cxgb4_main.c

/*
 * This file is part of the Chelsio T4 Ethernet driver for Linux.
 *
 * Copyright (c) 2003-2010 Chelsio Communications, Inc. All rights reserved.
 *
 * This software is available to you under a choice of one of two
 * licenses.  You may choose to be licensed under the terms of the GNU
 * General Public License (GPL) Version 2, available from the file
 * COPYING in the main directory of this source tree, or the
 * OpenIB.org BSD license below:
 *
 *     Redistribution and use in source and binary forms, with or
 *     without modification, are permitted provided that the following
 *     conditions are met:
 *
 *      - Redistributions of source code must retain the above
 *        copyright notice, this list of conditions and the following
 *        disclaimer.
 *
 *      - Redistributions in binary form must reproduce the above
 *        copyright notice, this list of conditions and the following
 *        disclaimer in the documentation and/or other materials
 *        provided with the distribution.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 */

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/bitmap.h>
#include <linux/crc32.h>
#include <linux/ctype.h>
#include <linux/debugfs.h>
#include <linux/err.h>
#include <linux/etherdevice.h>
#include <linux/firmware.h>
#include <linux/if.h>
#include <linux/if_vlan.h>
#include <linux/init.h>
#include <linux/log2.h>
#include <linux/mdio.h>
#include <linux/module.h>
#include <linux/moduleparam.h>
#include <linux/mutex.h>
#include <linux/netdevice.h>
#include <linux/pci.h>
#include <linux/aer.h>
#include <linux/rtnetlink.h>
#include <linux/sched.h>
#include <linux/seq_file.h>
#include <linux/sockios.h>
#include <linux/vmalloc.h>
#include <linux/workqueue.h>
#include <net/neighbour.h>
#include <net/netevent.h>
#include <asm/uaccess.h>

#include "cxgb4.h"
#include "t4_regs.h"
#include "t4_msg.h"
#include "t4fw_api.h"
#include "l2t.h"

#define DRV_VERSION "1.3.0-ko"
#define DRV_DESC "Chelsio T4 Network Driver"

/*
 * Max interrupt hold-off timer value in us.  Queues fall back to this value
 * under extreme memory pressure so it's largish to give the system time to
 * recover.
 */
#define MAX_SGE_TIMERVAL 200U

enum {
	/*
	 * Physical Function provisioning constants.
	 */
	PFRES_NVI = 4,			/* # of Virtual Interfaces */
	PFRES_NETHCTRL = 128,		/* # of EQs used for ETH or CTRL Qs */
	PFRES_NIQFLINT = 128,		/* # of ingress Qs/w Free List(s)/intr
					 */
	PFRES_NEQ = 256,		/* # of egress queues */
	PFRES_NIQ = 0,			/* # of ingress queues */
	PFRES_TC = 0,			/* PCI-E traffic class */
	PFRES_NEXACTF = 128,		/* # of exact MPS filters */

	PFRES_R_CAPS = FW_CMD_CAP_PF,
	PFRES_WX_CAPS = FW_CMD_CAP_PF,

#ifdef CONFIG_PCI_IOV
	/*
	 * Virtual Function provisioning constants.  We need two extra Ingress
	 * Queues with Interrupt capability to serve as the VF's Firmware
	 * Event Queue and Forwarded Interrupt Queue (when using MSI mode) --
	 * neither will have Free Lists associated with them).  For each
	 * Ethernet/Control Egress Queue and for each Free List, we need an
	 * Egress Context.
	 */
	VFRES_NPORTS = 1,		/* # of "ports" per VF */
	VFRES_NQSETS = 2,		/* # of "Queue Sets" per VF */

	VFRES_NVI = VFRES_NPORTS,	/* # of Virtual Interfaces */
	VFRES_NETHCTRL = VFRES_NQSETS,	/* # of EQs used for ETH or CTRL Qs */
	VFRES_NIQFLINT = VFRES_NQSETS+2,/* # of ingress Qs/w Free List(s)/intr */
	VFRES_NEQ = VFRES_NQSETS*2,	/* # of egress queues */
	VFRES_NIQ = 0,			/* # of non-fl/int ingress queues */
	VFRES_TC = 0,			/* PCI-E traffic class */
	VFRES_NEXACTF = 16,		/* # of exact MPS filters */

	VFRES_R_CAPS = FW_CMD_CAP_DMAQ|FW_CMD_CAP_VF|FW_CMD_CAP_PORT,
	VFRES_WX_CAPS = FW_CMD_CAP_DMAQ|FW_CMD_CAP_VF,
#endif
};

/*
 * Provide a Port Access Rights Mask for the specified PF/VF.  This is very
 * static and likely not to be useful in the long run.  We really need to
 * implement some form of persistent configuration which the firmware
 * controls.
 */
static unsigned int pfvfres_pmask(struct adapter *adapter,
				  unsigned int pf, unsigned int vf)
{
	unsigned int portn, portvec;

	/*
	 * Give PF's access to all of the ports.
	 */
	if (vf == 0)
		return FW_PFVF_CMD_PMASK_MASK;

	/*
	 * For VFs, we'll assign them access to the ports based purely on the
	 * PF.  We assign active ports in order, wrapping around if there are
	 * fewer active ports than PFs: e.g. active port[pf % nports].
	 * Unfortunately the adapter's port_info structs haven't been
	 * initialized yet so we have to compute this.
	 */
	if (adapter->params.nports == 0)
		return 0;

	portn = pf % adapter->params.nports;
	portvec = adapter->params.portvec;
	for (;;) {
		/*
		 * Isolate the lowest set bit in the port vector.  If we're at
		 * the port number that we want, return that as the pmask.
		 * otherwise mask that bit out of the port vector and
		 * decrement our port number ...
		 */
		unsigned int pmask = portvec ^ (portvec & (portvec-1));
		if (portn == 0)
			return pmask;
		portn--;
		portvec &= ~pmask;
	}
	/*NOTREACHED*/
}

enum {
	MAX_TXQ_ENTRIES      = 16384,
	MAX_CTRL_TXQ_ENTRIES = 1024,
	MAX_RSPQ_ENTRIES     = 16384,
	MAX_RX_BUFFERS       = 16384,
	MIN_TXQ_ENTRIES      = 32,
	MIN_CTRL_TXQ_ENTRIES = 32,
	MIN_RSPQ_ENTRIES     = 128,
	MIN_FL_ENTRIES       = 16
};

/* Host shadow copy of ingress filter entry.  This is in host native format
 * and doesn't match the ordering or bit order, etc. of the hardware of the
 * firmware command.  The use of bit-field structure elements is purely to
 * remind ourselves of the field size limitations and save memory in the case
 * where the filter table is large.
 */
struct filter_entry {
	/* Administrative fields for filter.
	 */
	u32 valid:1;            /* filter allocated and valid */
	u32 locked:1;           /* filter is administratively locked */

	u32 pending:1;          /* filter action is pending firmware reply */
	u32 smtidx:8;           /* Source MAC Table index for smac */
	struct l2t_entry *l2t;  /* Layer Two Table entry for dmac */

	/* The filter itself.  Most of this is a straight copy of information
	 * provided by the extended ioctl().  Some fields are translated to
	 * internal forms -- for instance the Ingress Queue ID passed in from
	 * the ioctl() is translated into the Absolute Ingress Queue ID.
	 */
	struct ch_filter_specification fs;
};

#define DFLT_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_PROBE | NETIF_MSG_LINK | \
			 NETIF_MSG_TIMER | NETIF_MSG_IFDOWN | NETIF_MSG_IFUP |\
			 NETIF_MSG_RX_ERR | NETIF_MSG_TX_ERR)

#define CH_DEVICE(devid, data) { PCI_VDEVICE(CHELSIO, devid), (data) }

static DEFINE_PCI_DEVICE_TABLE(cxgb4_pci_tbl) = {
	CH_DEVICE(0xa000, 0),  /* PE10K */
	CH_DEVICE(0x4001, -1),
	CH_DEVICE(0x4002, -1),
	CH_DEVICE(0x4003, -1),
	CH_DEVICE(0x4004, -1),
	CH_DEVICE(0x4005, -1),
	CH_DEVICE(0x4006, -1),
	CH_DEVICE(0x4007, -1),
	CH_DEVICE(0x4008, -1),
	CH_DEVICE(0x4009, -1),
	CH_DEVICE(0x400a, -1),
	CH_DEVICE(0x4401, 4),
	CH_DEVICE(0x4402, 4),
	CH_DEVICE(0x4403, 4),
	CH_DEVICE(0x4404, 4),
	CH_DEVICE(0x4405, 4),
	CH_DEVICE(0x4406, 4),
	CH_DEVICE(0x4407, 4),
	CH_DEVICE(0x4408, 4),
	CH_DEVICE(0x4409, 4),
	CH_DEVICE(0x440a, 4),
	CH_DEVICE(0x440d, 4),
	CH_DEVICE(0x440e, 4),
	{ 0, }
};

#define FW_FNAME "cxgb4/t4fw.bin"
#define FW_CFNAME "cxgb4/t4-config.txt"

MODULE_DESCRIPTION(DRV_DESC);
MODULE_AUTHOR("Chelsio Communications");
MODULE_LICENSE("Dual BSD/GPL");
MODULE_VERSION(DRV_VERSION);
MODULE_DEVICE_TABLE(pci, cxgb4_pci_tbl);
MODULE_FIRMWARE(FW_FNAME);

/*
 * Normally we're willing to become the firmware's Master PF but will be happy
 * if another PF has already become the Master and initialized the adapter.
 * Setting "force_init" will cause this driver to forcibly establish itself as
 * the Master PF and initialize the adapter.
 */
static uint force_init;

module_param(force_init, uint, 0644);
MODULE_PARM_DESC(force_init, "Forcibly become Master PF and initialize adapter");

/*
 * Normally if the firmware we connect to has Configuration File support, we
 * use that and only fall back to the old Driver-based initialization if the
 * Configuration File fails for some reason.  If force_old_init is set, then
 * we'll always use the old Driver-based initialization sequence.
 */
static uint force_old_init;

module_param(force_old_init, uint, 0644);
MODULE_PARM_DESC(force_old_init, "Force old initialization sequence");

static int dflt_msg_enable = DFLT_MSG_ENABLE;

module_param(dflt_msg_enable, int, 0644);
MODULE_PARM_DESC(dflt_msg_enable, "Chelsio T4 default message enable bitmap");

/*
 * The driver uses the best interrupt scheme available on a platform in the
 * order MSI-X, MSI, legacy INTx interrupts.  This parameter determines which
 * of these schemes the driver may consider as follows:
 *
 * msi = 2: choose from among all three options
 * msi = 1: only consider MSI and INTx interrupts
 * msi = 0: force INTx interrupts
 */
static int msi = 2;

module_param(msi, int, 0644);
MODULE_PARM_DESC(msi, "whether to use INTx (0), MSI (1) or MSI-X (2)");

/*
 * Queue interrupt hold-off timer values.  Queues default to the first of these
 * upon creation.
 */
static unsigned int intr_holdoff[SGE_NTIMERS - 1] = { 5, 10, 20, 50, 100 };

module_param_array(intr_holdoff, uint, NULL, 0644);
MODULE_PARM_DESC(intr_holdoff, "values for queue interrupt hold-off timers "
		 "0..4 in microseconds");

static unsigned int intr_cnt[SGE_NCOUNTERS - 1] = { 4, 8, 16 };

module_param_array(intr_cnt, uint, NULL, 0644);
MODULE_PARM_DESC(intr_cnt,
		 "thresholds 1..3 for queue interrupt packet counters");

/*
 * Normally we tell the chip to deliver Ingress Packets into our DMA buffers
 * offset by 2 bytes in order to have the IP headers line up on 4-byte
 * boundaries.  This is a requirement for many architectures which will throw
 * a machine check fault if an attempt is made to access one of the 4-byte IP
 * header fields on a non-4-byte boundary.  And it's a major performance issue
 * even on some architectures which allow it like some implementations of the
 * x86 ISA.  However, some architectures don't mind this and for some very
 * edge-case performance sensitive applications (like forwarding large volumes
 * of small packets), setting this DMA offset to 0 will decrease the number of
 * PCI-E Bus transfers enough to measurably affect performance.
 */
static int rx_dma_offset = 2;

static bool vf_acls;

#ifdef CONFIG_PCI_IOV
module_param(vf_acls, bool, 0644);
MODULE_PARM_DESC(vf_acls, "if set enable virtualization L2 ACL enforcement");

static unsigned int num_vf[4];

module_param_array(num_vf, uint, NULL, 0644);
MODULE_PARM_DESC(num_vf, "number of VFs for each of PFs 0-3");
#endif

/*
 * The filter TCAM has a fixed portion and a variable portion.  The fixed
 * portion can match on source/destination IP IPv4/IPv6 addresses and TCP/UDP
 * ports.  The variable portion is 36 bits which can include things like Exact
 * Match MAC Index (9 bits), Ether Type (16 bits), IP Protocol (8 bits),
 * [Inner] VLAN Tag (17 bits), etc. which, if all were somehow selected, would
 * far exceed the 36-bit budget for this "compressed" header portion of the
 * filter.  Thus, we have a scarce resource which must be carefully managed.
 *
 * By default we set this up to mostly match the set of filter matching
 * capabilities of T3 but with accommodations for some of T4's more
 * interesting features:
 *
 *   { IP Fragment (1), MPS Match Type (3), IP Protocol (8),
 *     [Inner] VLAN (17), Port (3), FCoE (1) }
 */
enum {
	TP_VLAN_PRI_MAP_DEFAULT = HW_TPL_FR_MT_PR_IV_P_FC,
	TP_VLAN_PRI_MAP_FIRST = FCOE_SHIFT,
	TP_VLAN_PRI_MAP_LAST = FRAGMENTATION_SHIFT,
};

static unsigned int tp_vlan_pri_map = TP_VLAN_PRI_MAP_DEFAULT;

module_param(tp_vlan_pri_map, uint, 0644);
MODULE_PARM_DESC(tp_vlan_pri_map, "global compressed filter configuration");

static struct dentry *cxgb4_debugfs_root;

static LIST_HEAD(adapter_list);
static DEFINE_MUTEX(uld_mutex);
static struct cxgb4_uld_info ulds[CXGB4_ULD_MAX];
static const char *uld_str[] = { "RDMA", "iSCSI" };

static void link_report(struct net_device *dev)
{
	if (!netif_carrier_ok(dev))
		netdev_info(dev, "link down\n");
	else {
		static const char *fc[] = { "no", "Rx", "Tx", "Tx/Rx" };

		const char *s = "10Mbps";
		const struct port_info *p = netdev_priv(dev);

		switch (p->link_cfg.speed) {
		case SPEED_10000:
			s = "10Gbps";
			break;
		case SPEED_1000:
			s = "1000Mbps";
			break;
		case SPEED_100:
			s = "100Mbps";
			break;
		}

		netdev_info(dev, "link up, %s, full-duplex, %s PAUSE\n", s,
			    fc[p->link_cfg.fc]);
	}
}

void t4_os_link_changed(struct adapter *adapter, int port_id, int link_stat)
{
	struct net_device *dev = adapter->port[port_id];

	/* Skip changes from disabled ports. */
	if (netif_running(dev) && link_stat != netif_carrier_ok(dev)) {
		if (link_stat)
			netif_carrier_on(dev);
		else
			netif_carrier_off(dev);

		link_report(dev);
	}
}

void t4_os_portmod_changed(const struct adapter *adap, int port_id)
{
	static const char *mod_str[] = {
		NULL, "LR", "SR", "ER", "passive DA", "active DA", "LRM"
	};

	const struct net_device *dev = adap->port[port_id];
	const struct port_info *pi = netdev_priv(dev);

	if (pi->mod_type == FW_PORT_MOD_TYPE_NONE)
		netdev_info(dev, "port module unplugged\n");
	else if (pi->mod_type < ARRAY_SIZE(mod_str))
		netdev_info(dev, "%s module inserted\n", mod_str[pi->mod_type]);
}

/*
 * Configure the exact and hash address filters to handle a port's multicast
 * and secondary unicast MAC addresses.
 */
static int set_addr_filters(const struct net_device *dev, bool sleep)
{
	u64 mhash = 0;
	u64 uhash = 0;
	bool free = true;
	u16 filt_idx[7];
	const u8 *addr[7];
	int ret, naddr = 0;
	const struct netdev_hw_addr *ha;
	int uc_cnt = netdev_uc_count(dev);
	int mc_cnt = netdev_mc_count(dev);
	const struct port_info *pi = netdev_priv(dev);
	unsigned int mb = pi->adapter->fn;

	/* first do the secondary unicast addresses */
	netdev_for_each_uc_addr(ha, dev) {
		addr[naddr++] = ha->addr;
		if (--uc_cnt == 0 || naddr >= ARRAY_SIZE(addr)) {
			ret = t4_alloc_mac_filt(pi->adapter, mb, pi->viid, free,
					naddr, addr, filt_idx, &uhash, sleep);
			if (ret < 0)
				return ret;

			free = false;
			naddr = 0;
		}
	}

	/* next set up the multicast addresses */
	netdev_for_each_mc_addr(ha, dev) {
		addr[naddr++] = ha->addr;
		if (--mc_cnt == 0 || naddr >= ARRAY_SIZE(addr)) {
			ret = t4_alloc_mac_filt(pi->adapter, mb, pi->viid, free,
					naddr, addr, filt_idx, &mhash, sleep);
			if (ret < 0)
				return ret;

			free = false;
			naddr = 0;
		}
	}

	return t4_set_addr_hash(pi->adapter, mb, pi->viid, uhash != 0,
				uhash | mhash, sleep);
}

int dbfifo_int_thresh = 10; /* 10 == 640 entry threshold */
module_param(dbfifo_int_thresh, int, 0644);
MODULE_PARM_DESC(dbfifo_int_thresh, "doorbell fifo interrupt threshold");

/*
 * usecs to sleep while draining the dbfifo
 */
static int dbfifo_drain_delay = 1000;
module_param(dbfifo_drain_delay, int, 0644);
MODULE_PARM_DESC(dbfifo_drain_delay,
		 "usecs to sleep while draining the dbfifo");

/*
 * Set Rx properties of a port, such as promiscruity, address filters, and MTU.
 * If @mtu is -1 it is left unchanged.
 */
static int set_rxmode(struct net_device *dev, int mtu, bool sleep_ok)
{
	int ret;
	struct port_info *pi = netdev_priv(dev);

	ret = set_addr_filters(dev, sleep_ok);
	if (ret == 0)
		ret = t4_set_rxmode(pi->adapter, pi->adapter->fn, pi->viid, mtu,
				    (dev->flags & IFF_PROMISC) ? 1 : 0,
				    (dev->flags & IFF_ALLMULTI) ? 1 : 0, 1, -1,
				    sleep_ok);
	return ret;
}

static struct workqueue_struct *workq;

/**
 *	link_start - enable a port
 *	@dev: the port to enable
 *
 *	Performs the MAC and PHY actions needed to enable a port.
 */
static int link_start(struct net_device *dev)
{
	int ret;
	struct port_info *pi = netdev_priv(dev);
	unsigned int mb = pi->adapter->fn;

	/*
	 * We do not set address filters and promiscuity here, the stack does
	 * that step explicitly.
	 */
	ret = t4_set_rxmode(pi->adapter, mb, pi->viid, dev->mtu, -1, -1, -1,
			    !!(dev->features & NETIF_F_HW_VLAN_RX), true);
	if (ret == 0) {
		ret = t4_change_mac(pi->adapter, mb, pi->viid,
				    pi->xact_addr_filt, dev->dev_addr, true,
				    true);
		if (ret >= 0) {
			pi->xact_addr_filt = ret;
			ret = 0;
		}
	}
	if (ret == 0)
		ret = t4_link_start(pi->adapter, mb, pi->tx_chan,
				    &pi->link_cfg);
	if (ret == 0)
		ret = t4_enable_vi(pi->adapter, mb, pi->viid, true, true);
	return ret;
}

/* Clear a filter and release any of its resources that we own.  This also
 * clears the filter's "pending" status.
 */
static void clear_filter(struct adapter *adap, struct filter_entry *f)
{
	/* If the new or old filter have loopback rewriteing rules then we'll
	 * need to free any existing Layer Two Table (L2T) entries of the old
	 * filter rule.  The firmware will handle freeing up any Source MAC
	 * Table (SMT) entries used for rewriting Source MAC Addresses in
	 * loopback rules.
	 */
	if (f->l2t)
		cxgb4_l2t_release(f->l2t);

	/* The zeroing of the filter rule below clears the filter valid,
	 * pending, locked flags, l2t pointer, etc. so it's all we need for
	 * this operation.
	 */
	memset(f, 0, sizeof(*f));
}

/* Handle a filter write/deletion reply.
 */
static void filter_rpl(struct adapter *adap, const struct cpl_set_tcb_rpl *rpl)
{
	unsigned int idx = GET_TID(rpl);
	unsigned int nidx = idx - adap->tids.ftid_base;
	unsigned int ret;
	struct filter_entry *f;

	if (idx >= adap->tids.ftid_base && nidx <
	   (adap->tids.nftids + adap->tids.nsftids)) {
		idx = nidx;
		ret = GET_TCB_COOKIE(rpl->cookie);
		f = &adap->tids.ftid_tab[idx];

		if (ret == FW_FILTER_WR_FLT_DELETED) {
			/* Clear the filter when we get confirmation from the
			 * hardware that the filter has been deleted.
			 */
			clear_filter(adap, f);
		} else if (ret == FW_FILTER_WR_SMT_TBL_FULL) {
			dev_err(adap->pdev_dev, "filter %u setup failed due to full SMT\n",
				idx);
			clear_filter(adap, f);
		} else if (ret == FW_FILTER_WR_FLT_ADDED) {
			f->smtidx = (be64_to_cpu(rpl->oldval) >> 24) & 0xff;
			f->pending = 0;  /* asynchronous setup completed */
			f->valid = 1;
		} else {
			/* Something went wrong.  Issue a warning about the
			 * problem and clear everything out.
			 */
			dev_err(adap->pdev_dev, "filter %u setup failed with error %u\n",
				idx, ret);
			clear_filter(adap, f);
		}
	}
}

/* Response queue handler for the FW event queue.
 */
static int fwevtq_handler(struct sge_rspq *q, const __be64 *rsp,
			  const struct pkt_gl *gl)
{
	u8 opcode = ((const struct rss_header *)rsp)->opcode;

	rsp++;                                          /* skip RSS header */
	if (likely(opcode == CPL_SGE_EGR_UPDATE)) {
		const struct cpl_sge_egr_update *p = (void *)rsp;
		unsigned int qid = EGR_QID(ntohl(p->opcode_qid));
		struct sge_txq *txq;

		txq = q->adap->sge.egr_map[qid - q->adap->sge.egr_start];
		txq->restarts++;
		if ((u8 *)txq < (u8 *)q->adap->sge.ofldtxq) {
			struct sge_eth_txq *eq;

			eq = container_of(txq, struct sge_eth_txq, q);
			netif_tx_wake_queue(eq->txq);
		} else {
			struct sge_ofld_txq *oq;

			oq = container_of(txq, struct sge_ofld_txq, q);
			tasklet_schedule(&oq->qresume_tsk);
		}
	} else if (opcode == CPL_FW6_MSG || opcode == CPL_FW4_MSG) {
		const struct cpl_fw6_msg *p = (void *)rsp;

		if (p->type == 0)
			t4_handle_fw_rpl(q->adap, p->data);
	} else if (opcode == CPL_L2T_WRITE_RPL) {
		const struct cpl_l2t_write_rpl *p = (void *)rsp;

		do_l2t_write_rpl(q->adap, p);
	} else if (opcode == CPL_SET_TCB_RPL) {
		const struct cpl_set_tcb_rpl *p = (void *)rsp;

		filter_rpl(q->adap, p);
	} else
		dev_err(q->adap->pdev_dev,
			"unexpected CPL %#x on FW event queue\n", opcode);
	return 0;
}

/**
 *	uldrx_handler - response queue handler for ULD queues
 *	@q: the response queue that received the packet
 *	@rsp: the response queue descriptor holding the offload message
 *	@gl: the gather list of packet fragments
 *
 *	Deliver an ingress offload packet to a ULD.  All processing is done by
 *	the ULD, we just maintain statistics.
 */
static int uldrx_handler(struct sge_rspq *q, const __be64 *rsp,
			 const struct pkt_gl *gl)
{
	struct sge_ofld_rxq *rxq = container_of(q, struct sge_ofld_rxq, rspq);

	if (ulds[q->uld].rx_handler(q->adap->uld_handle[q->uld], rsp, gl)) {
		rxq->stats.nomem++;
		return -1;
	}
	if (gl == NULL)
		rxq->stats.imm++;
	else if (gl == CXGB4_MSG_AN)
		rxq->stats.an++;
	else
		rxq->stats.pkts++;
	return 0;
}

static void disable_msi(struct adapter *adapter)
{
	if (adapter->flags & USING_MSIX) {
		pci_disable_msix(adapter->pdev);
		adapter->flags &= ~USING_MSIX;
	} else if (adapter->flags & USING_MSI) {
		pci_disable_msi(adapter->pdev);
		adapter->flags &= ~USING_MSI;
	}
}

/*
 * Interrupt handler for non-data events used with MSI-X.
 */
static irqreturn_t t4_nondata_intr(int irq, void *cookie)
{
	struct adapter *adap = cookie;

	u32 v = t4_read_reg(adap, MYPF_REG(PL_PF_INT_CAUSE));
	if (v & PFSW) {
		adap->swintr = 1;
		t4_write_reg(adap, MYPF_REG(PL_PF_INT_CAUSE), v);
	}
	t4_slow_intr_handler(adap);
	return IRQ_HANDLED;
}

/*
 * Name the MSI-X interrupts.
 */
static void name_msix_vecs(struct adapter *adap)
{
	int i, j, msi_idx = 2, n = sizeof(adap->msix_info[0].desc);

	/* non-data interrupts */
	snprintf(adap->msix_info[0].desc, n, "%s", adap->port[0]->name);

	/* FW events */
	snprintf(adap->msix_info[1].desc, n, "%s-FWeventq",
		 adap->port[0]->name);

	/* Ethernet queues */
	for_each_port(adap, j) {
		struct net_device *d = adap->port[j];
		const struct port_info *pi = netdev_priv(d);

		for (i = 0; i < pi->nqsets; i++, msi_idx++)
			snprintf(adap->msix_info[msi_idx].desc, n, "%s-Rx%d",
				 d->name, i);
	}

	/* offload queues */
	for_each_ofldrxq(&adap->sge, i)
		snprintf(adap->msix_info[msi_idx++].desc, n, "%s-ofld%d",
			 adap->port[0]->name, i);

	for_each_rdmarxq(&adap->sge, i)
		snprintf(adap->msix_info[msi_idx++].desc, n, "%s-rdma%d",
			 adap->port[0]->name, i);
}

static int request_msix_queue_irqs(struct adapter *adap)
{
	struct sge *s = &adap->sge;
	int err, ethqidx, ofldqidx = 0, rdmaqidx = 0, msi_index = 2;

	err = request_irq(adap->msix_info[1].vec, t4_sge_intr_msix, 0,
			  adap->msix_info[1].desc, &s->fw_evtq);
	if (err)
		return err;

	for_each_ethrxq(s, ethqidx) {
		err = request_irq(adap->msix_info[msi_index].vec,
				  t4_sge_intr_msix, 0,
				  adap->msix_info[msi_index].desc,
				  &s->ethrxq[ethqidx].rspq);
		if (err)
			goto unwind;
		msi_index++;
	}
	for_each_ofldrxq(s, ofldqidx) {
		err = request_irq(adap->msix_info[msi_index].vec,
				  t4_sge_intr_msix, 0,
				  adap->msix_info[msi_index].desc,
				  &s->ofldrxq[ofldqidx].rspq);
		if (err)
			goto unwind;
		msi_index++;
	}
	for_each_rdmarxq(s, rdmaqidx) {
		err = request_irq(adap->msix_info[msi_index].vec,
				  t4_sge_intr_msix, 0,
				  adap->msix_info[msi_index].desc,
				  &s->rdmarxq[rdmaqidx].rspq);
		if (err)
			goto unwind;
		msi_index++;
	}
	return 0;

unwind:
	while (--rdmaqidx >= 0)
		free_irq(adap->msix_info[--msi_index].vec,
			 &s->rdmarxq[rdmaqidx].rspq);
	while (--ofldqidx >= 0)
		free_irq(adap->msix_info[--msi_index].vec,
			 &s->ofldrxq[ofldqidx].rspq);
	while (--ethqidx >= 0)
		free_irq(adap->msix_info[--msi_index].vec,
			 &s->ethrxq[ethqidx].rspq);
	free_irq(adap->msix_info[1].vec, &s->fw_evtq);
	return err;
}

static void free_msix_queue_irqs(struct adapter *adap)
{
	int i, msi_index = 2;
	struct sge *s = &adap->sge;

	free_irq(adap->msix_info[1].vec, &s->fw_evtq);
	for_each_ethrxq(s, i)
		free_irq(adap->msix_info[msi_index++].vec, &s->ethrxq[i].rspq);
	for_each_ofldrxq(s, i)
		free_irq(adap->msix_info[msi_index++].vec, &s->ofldrxq[i].rspq);
	for_each_rdmarxq(s, i)
		free_irq(adap->msix_info[msi_index++].vec, &s->rdmarxq[i].rspq);
}

/**
 *	write_rss - write the RSS table for a given port
 *	@pi: the port
 *	@queues: array of queue indices for RSS
 *
 *	Sets up the portion of the HW RSS table for the port's VI to distribute
 *	packets to the Rx queues in @queues.
 */
static int write_rss(const struct port_info *pi, const u16 *queues)
{
	u16 *rss;
	int i, err;
	const struct sge_eth_rxq *q = &pi->adapter->sge.ethrxq[pi->first_qset];

	rss = kmalloc(pi->rss_size * sizeof(u16), GFP_KERNEL);
	if (!rss)
		return -ENOMEM;

	/* map the queue indices to queue ids */
	for (i = 0; i < pi->rss_size; i++, queues++)
		rss[i] = q[*queues].rspq.abs_id;

	err = t4_config_rss_range(pi->adapter, pi->adapter->fn, pi->viid, 0,
				  pi->rss_size, rss, pi->rss_size);
	kfree(rss);
	return err;
}

/**
 *	setup_rss - configure RSS
 *	@adap: the adapter
 *
 *	Sets up RSS for each port.
 */
static int setup_rss(struct adapter *adap)
{
	int i, err;

	for_each_port(adap, i) {
		const struct port_info *pi = adap2pinfo(adap, i);

		err = write_rss(pi, pi->rss);
		if (err)
			return err;
	}
	return 0;
}

/*
 * Return the channel of the ingress queue with the given qid.
 */
static unsigned int rxq_to_chan(const struct sge *p, unsigned int qid)
{
	qid -= p->ingr_start;
	return netdev2pinfo(p->ingr_map[qid]->netdev)->tx_chan;
}

/*
 * Wait until all NAPI handlers are descheduled.
 */
static void quiesce_rx(struct adapter *adap)
{
	int i;

	for (i = 0; i < ARRAY_SIZE(adap->sge.ingr_map); i++) {
		struct sge_rspq *q = adap->sge.ingr_map[i];

		if (q && q->handler)
			napi_disable(&q->napi);
	}
}

/*
 * Enable NAPI scheduling and interrupt generation for all Rx queues.
 */
static void enable_rx(struct adapter *adap)
{
	int i;

	for (i = 0; i < ARRAY_SIZE(adap->sge.ingr_map); i++) {
		struct sge_rspq *q = adap->sge.ingr_map[i];

		if (!q)
			continue;
		if (q->handler)
			napi_enable(&q->napi);
		/* 0-increment GTS to start the timer and enable interrupts */
		t4_write_reg(adap, MYPF_REG(SGE_PF_GTS),
			     SEINTARM(q->intr_params) |
			     INGRESSQID(q->cntxt_id));
	}
}

/**
 *	setup_sge_queues - configure SGE Tx/Rx/response queues
 *	@adap: the adapter
 *
 *	Determines how many sets of SGE queues to use and initializes them.
 *	We support multiple queue sets per port if we have MSI-X, otherwise
 *	just one queue set per port.
 */
static int setup_sge_queues(struct adapter *adap)
{
	int err, msi_idx, i, j;
	struct sge *s = &adap->sge;

	bitmap_zero(s->starving_fl, MAX_EGRQ);
	bitmap_zero(s->txq_maperr, MAX_EGRQ);

	if (adap->flags & USING_MSIX)
		msi_idx = 1;         /* vector 0 is for non-queue interrupts */
	else {
		err = t4_sge_alloc_rxq(adap, &s->intrq, false, adap->port[0], 0,
				       NULL, NULL);
		if (err)
			return err;
		msi_idx = -((int)s->intrq.abs_id + 1);
	}

	err = t4_sge_alloc_rxq(adap, &s->fw_evtq, true, adap->port[0],
			       msi_idx, NULL, fwevtq_handler);
	if (err) {
freeout:	t4_free_sge_resources(adap);
		return err;
	}

	for_each_port(adap, i) {
		struct net_device *dev = adap->port[i];
		struct port_info *pi = netdev_priv(dev);
		struct sge_eth_rxq *q = &s->ethrxq[pi->first_qset];
		struct sge_eth_txq *t = &s->ethtxq[pi->first_qset];

		for (j = 0; j < pi->nqsets; j++, q++) {
			if (msi_idx > 0)
				msi_idx++;
			err = t4_sge_alloc_rxq(adap, &q->rspq, false, dev,
					       msi_idx, &q->fl,
					       t4_ethrx_handler);
			if (err)
				goto freeout;
			q->rspq.idx = j;
			memset(&q->stats, 0, sizeof(q->stats));
		}
		for (j = 0; j < pi->nqsets; j++, t++) {
			err = t4_sge_alloc_eth_txq(adap, t, dev,
					netdev_get_tx_queue(dev, j),
					s->fw_evtq.cntxt_id);
			if (err)
				goto freeout;
		}
	}

	j = s->ofldqsets / adap->params.nports; /* ofld queues per channel */
	for_each_ofldrxq(s, i) {
		struct sge_ofld_rxq *q = &s->ofldrxq[i];
		struct net_device *dev = adap->port[i / j];

		if (msi_idx > 0)
			msi_idx++;
		err = t4_sge_alloc_rxq(adap, &q->rspq, false, dev, msi_idx,
				       &q->fl, uldrx_handler);
		if (err)
			goto freeout;
		memset(&q->stats, 0, sizeof(q->stats));
		s->ofld_rxq[i] = q->rspq.abs_id;
		err = t4_sge_alloc_ofld_txq(adap, &s->ofldtxq[i], dev,
					    s->fw_evtq.cntxt_id);
		if (err)
			goto freeout;
	}

	for_each_rdmarxq(s, i) {
		struct sge_ofld_rxq *q = &s->rdmarxq[i];

		if (msi_idx > 0)
			msi_idx++;
		err = t4_sge_alloc_rxq(adap, &q->rspq, false, adap->port[i],
				       msi_idx, &q->fl, uldrx_handler);
		if (err)
			goto freeout;
		memset(&q->stats, 0, sizeof(q->stats));
		s->rdma_rxq[i] = q->rspq.abs_id;
	}

	for_each_port(adap, i) {
		/*
		 * Note that ->rdmarxq[i].rspq.cntxt_id below is 0 if we don't
		 * have RDMA queues, and that's the right value.
		 */
		err = t4_sge_alloc_ctrl_txq(adap, &s->ctrlq[i], adap->port[i],
					    s->fw_evtq.cntxt_id,
					    s->rdmarxq[i].rspq.cntxt_id);
		if (err)
			goto freeout;
	}

	t4_write_reg(adap, MPS_TRC_RSS_CONTROL,
		     RSSCONTROL(netdev2pinfo(adap->port[0])->tx_chan) |
		     QUEUENUMBER(s->ethrxq[0].rspq.abs_id));
	return 0;
}

/*
 * Returns 0 if new FW was successfully loaded, a positive errno if a load was
 * started but failed, and a negative errno if flash load couldn't start.
 */
static int upgrade_fw(struct adapter *adap)
{
	int ret;
	u32 vers;
	const struct fw_hdr *hdr;
	const struct firmware *fw;
	struct device *dev = adap->pdev_dev;

	ret = request_firmware(&fw, FW_FNAME, dev);
	if (ret < 0) {
		dev_err(dev, "unable to load firmware image " FW_FNAME
			", error %d\n", ret);
		return ret;
	}

	hdr = (const struct fw_hdr *)fw->data;
	vers = ntohl(hdr->fw_ver);
	if (FW_HDR_FW_VER_MAJOR_GET(vers) != FW_VERSION_MAJOR) {
		ret = -EINVAL;              /* wrong major version, won't do */
		goto out;
	}

	/*
	 * If the flash FW is unusable or we found something newer, load it.
	 */
	if (FW_HDR_FW_VER_MAJOR_GET(adap->params.fw_vers) != FW_VERSION_MAJOR ||
	    vers > adap->params.fw_vers) {
		dev_info(dev, "upgrading firmware ...\n");
		ret = t4_fw_upgrade(adap, adap->mbox, fw->data, fw->size,
				    /*force=*/false);
		if (!ret)
			dev_info(dev, "firmware successfully upgraded to "
				 FW_FNAME " (%d.%d.%d.%d)\n",
				 FW_HDR_FW_VER_MAJOR_GET(vers),
				 FW_HDR_FW_VER_MINOR_GET(vers),
				 FW_HDR_FW_VER_MICRO_GET(vers),
				 FW_HDR_FW_VER_BUILD_GET(vers));
		else
			dev_err(dev, "firmware upgrade failed! err=%d\n", -ret);
	} else {
		/*
		 * Tell our caller that we didn't upgrade the firmware.
		 */
		ret = -EINVAL;
	}

out:	release_firmware(fw);
	return ret;
}

/*
 * Allocate a chunk of memory using kmalloc or, if that fails, vmalloc.
 * The allocated memory is cleared.
 */
void *t4_alloc_mem(size_t size)
{
	void *p = kzalloc(size, GFP_KERNEL);

	if (!p)
		p = vzalloc(size);
	return p;
}

/*
 * Free memory allocated through alloc_mem().
 */
static void t4_free_mem(void *addr)
{
	if (is_vmalloc_addr(addr))
		vfree(addr);
	else
		kfree(addr);
}

/* Send a Work Request to write the filter at a specified index.  We construct
 * a Firmware Filter Work Request to have the work done and put the indicated
 * filter into "pending" mode which will prevent any further actions against
 * it till we get a reply from the firmware on the completion status of the
 * request.
 */
static int set_filter_wr(struct adapter *adapter, int fidx)
{
	struct filter_entry *f = &adapter->tids.ftid_tab[fidx];
	struct sk_buff *skb;
	struct fw_filter_wr *fwr;
	unsigned int ftid;

	/* If the new filter requires loopback Destination MAC and/or VLAN
	 * rewriting then we need to allocate a Layer 2 Table (L2T) entry for
	 * the filter.
	 */
	if (f->fs.newdmac || f->fs.newvlan) {
		/* allocate L2T entry for new filter */
		f->l2t = t4_l2t_alloc_switching(adapter->l2t);
		if (f->l2t == NULL)
			return -EAGAIN;
		if (t4_l2t_set_switching(adapter, f->l2t, f->fs.vlan,
					f->fs.eport, f->fs.dmac)) {
			cxgb4_l2t_release(f->l2t);
			f->l2t = NULL;
			return -ENOMEM;
		}
	}

	ftid = adapter->tids.ftid_base + fidx;

	skb = alloc_skb(sizeof(*fwr), GFP_KERNEL | __GFP_NOFAIL);
	fwr = (struct fw_filter_wr *)__skb_put(skb, sizeof(*fwr));
	memset(fwr, 0, sizeof(*fwr));

	/* It would be nice to put most of the following in t4_hw.c but most
	 * of the work is translating the cxgbtool ch_filter_specification
	 * into the Work Request and the definition of that structure is
	 * currently in cxgbtool.h which isn't appropriate to pull into the
	 * common code.  We may eventually try to come up with a more neutral
	 * filter specification structure but for now it's easiest to simply
	 * put this fairly direct code in line ...
	 */
	fwr->op_pkd = htonl(FW_WR_OP(FW_FILTER_WR));
	fwr->len16_pkd = htonl(FW_WR_LEN16(sizeof(*fwr)/16));
	fwr->tid_to_iq =
		htonl(V_FW_FILTER_WR_TID(ftid) |
		      V_FW_FILTER_WR_RQTYPE(f->fs.type) |
		      V_FW_FILTER_WR_NOREPLY(0) |
		      V_FW_FILTER_WR_IQ(f->fs.iq));
	fwr->del_filter_to_l2tix =
		htonl(V_FW_FILTER_WR_RPTTID(f->fs.rpttid) |
		      V_FW_FILTER_WR_DROP(f->fs.action == FILTER_DROP) |
		      V_FW_FILTER_WR_DIRSTEER(f->fs.dirsteer) |
		      V_FW_FILTER_WR_MASKHASH(f->fs.maskhash) |
		      V_FW_FILTER_WR_DIRSTEERHASH(f->fs.dirsteerhash) |
		      V_FW_FILTER_WR_LPBK(f->fs.action == FILTER_SWITCH) |
		      V_FW_FILTER_WR_DMAC(f->fs.newdmac) |
		      V_FW_FILTER_WR_SMAC(f->fs.newsmac) |
		      V_FW_FILTER_WR_INSVLAN(f->fs.newvlan == VLAN_INSERT ||
					     f->fs.newvlan == VLAN_REWRITE) |
		      V_FW_FILTER_WR_RMVLAN(f->fs.newvlan == VLAN_REMOVE ||
					    f->fs.newvlan == VLAN_REWRITE) |
		      V_FW_FILTER_WR_HITCNTS(f->fs.hitcnts) |
		      V_FW_FILTER_WR_TXCHAN(f->fs.eport) |
		      V_FW_FILTER_WR_PRIO(f->fs.prio) |
		      V_FW_FILTER_WR_L2TIX(f->l2t ? f->l2t->idx : 0));
	fwr->ethtype = htons(f->fs.val.ethtype);
	fwr->ethtypem = htons(f->fs.mask.ethtype);
	fwr->frag_to_ovlan_vldm =
		(V_FW_FILTER_WR_FRAG(f->fs.val.frag) |
		 V_FW_FILTER_WR_FRAGM(f->fs.mask.frag) |
		 V_FW_FILTER_WR_IVLAN_VLD(f->fs.val.ivlan_vld) |
		 V_FW_FILTER_WR_OVLAN_VLD(f->fs.val.ovlan_vld) |
		 V_FW_FILTER_WR_IVLAN_VLDM(f->fs.mask.ivlan_vld) |
		 V_FW_FILTER_WR_OVLAN_VLDM(f->fs.mask.ovlan_vld));
	fwr->smac_sel = 0;
	fwr->rx_chan_rx_rpl_iq =
		htons(V_FW_FILTER_WR_RX_CHAN(0) |
		      V_FW_FILTER_WR_RX_RPL_IQ(adapter->sge.fw_evtq.abs_id));
	fwr->maci_to_matchtypem =
		htonl(V_FW_FILTER_WR_MACI(f->fs.val.macidx) |
		      V_FW_FILTER_WR_MACIM(f->fs.mask.macidx) |
		      V_FW_FILTER_WR_FCOE(f->fs.val.fcoe) |
		      V_FW_FILTER_WR_FCOEM(f->fs.mask.fcoe) |
		      V_FW_FILTER_WR_PORT(f->fs.val.iport) |
		      V_FW_FILTER_WR_PORTM(f->fs.mask.iport) |
		      V_FW_FILTER_WR_MATCHTYPE(f->fs.val.matchtype) |
		      V_FW_FILTER_WR_MATCHTYPEM(f->fs.mask.matchtype));
	fwr->ptcl = f->fs.val.proto;
	fwr->ptclm = f->fs.mask.proto;
	fwr->ttyp = f->fs.val.tos;
	fwr->ttypm = f->fs.mask.tos;
	fwr->ivlan = htons(f->fs.val.ivlan);
	fwr->ivlanm = htons(f->fs.mask.ivlan);
	fwr->ovlan = htons(f->fs.val.ovlan);
	fwr->ovlanm = htons(f->fs.mask.ovlan);
	memcpy(fwr->lip, f->fs.val.lip, sizeof(fwr->lip));
	memcpy(fwr->lipm, f->fs.mask.lip, sizeof(fwr->lipm));
	memcpy(fwr->fip, f->fs.val.fip, sizeof(fwr->fip));
	memcpy(fwr->fipm, f->fs.mask.fip, sizeof(fwr->fipm));
	fwr->lp = htons(f->fs.val.lport);
	fwr->lpm = htons(f->fs.mask.lport);
	fwr->fp = htons(f->fs.val.fport);
	fwr->fpm = htons(f->fs.mask.fport);
	if (f->fs.newsmac)
		memcpy(fwr->sma, f->fs.smac, sizeof(fwr->sma));

	/* Mark the filter as "pending" and ship off the Filter Work Request.
	 * When we get the Work Request Reply we'll clear the pending status.
	 */
	f->pending = 1;
	set_wr_txq(skb, CPL_PRIORITY_CONTROL, f->fs.val.iport & 0x3);
	t4_ofld_send(adapter, skb);
	return 0;
}

/* Delete the filter at a specified index.
 */
static int del_filter_wr(struct adapter *adapter, int fidx)
{
	struct filter_entry *f = &adapter->tids.ftid_tab[fidx];
	struct sk_buff *skb;
	struct fw_filter_wr *fwr;
	unsigned int len, ftid;

	len = sizeof(*fwr);
	ftid = adapter->tids.ftid_base + fidx;

	skb = alloc_skb(len, GFP_KERNEL | __GFP_NOFAIL);
	fwr = (struct fw_filter_wr *)__skb_put(skb, len);
	t4_mk_filtdelwr(ftid, fwr, adapter->sge.fw_evtq.abs_id);

	/* Mark the filter as "pending" and ship off the Filter Work Request.
	 * When we get the Work Request Reply we'll clear the pending status.
	 */
	f->pending = 1;
	t4_mgmt_tx(adapter, skb);
	return 0;
}

static inline int is_offload(const struct adapter *adap)
{
	return adap->params.offload;
}

/*
 * Implementation of ethtool operations.
 */

static u32 get_msglevel(struct net_device *dev)
{
	return netdev2adap(dev)->msg_enable;
}

static void set_msglevel(struct net_device *dev, u32 val)
{
	netdev2adap(dev)->msg_enable = val;
}

static char stats_strings[][ETH_GSTRING_LEN] = {
	"TxOctetsOK         ",
	"TxFramesOK         ",
	"TxBroadcastFrames  ",
	"TxMulticastFrames  ",
	"TxUnicastFrames    ",
	"TxErrorFrames      ",

	"TxFrames64         ",
	"TxFrames65To127    ",
	"TxFrames128To255   ",
	"TxFrames256To511   ",
	"TxFrames512To1023  ",
	"TxFrames1024To1518 ",
	"TxFrames1519ToMax  ",

	"TxFramesDropped    ",
	"TxPauseFrames      ",
	"TxPPP0Frames       ",
	"TxPPP1Frames       ",
	"TxPPP2Frames       ",
	"TxPPP3Frames       ",
	"TxPPP4Frames       ",
	"TxPPP5Frames       ",
	"TxPPP6Frames       ",
	"TxPPP7Frames       ",

	"RxOctetsOK         ",
	"RxFramesOK         ",
	"RxBroadcastFrames  ",
	"RxMulticastFrames  ",
	"RxUnicastFrames    ",

	"RxFramesTooLong    ",
	"RxJabberErrors     ",
	"RxFCSErrors        ",
	"RxLengthErrors     ",
	"RxSymbolErrors     ",
	"RxRuntFrames       ",

	"RxFrames64         ",
	"RxFrames65To127    ",
	"RxFrames128To255   ",
	"RxFrames256To511   ",
	"RxFrames512To1023  ",
	"RxFrames1024To1518 ",
	"RxFrames1519ToMax  ",

	"RxPauseFrames      ",
	"RxPPP0Frames       ",
	"RxPPP1Frames       ",
	"RxPPP2Frames       ",
	"RxPPP3Frames       ",
	"RxPPP4Frames       ",
	"RxPPP5Frames       ",
	"RxPPP6Frames       ",
	"RxPPP7Frames       ",

	"RxBG0FramesDropped ",
	"RxBG1FramesDropped ",
	"RxBG2FramesDropped ",
	"RxBG3FramesDropped ",
	"RxBG0FramesTrunc   ",
	"RxBG1FramesTrunc   ",
	"RxBG2FramesTrunc   ",
	"RxBG3FramesTrunc   ",

	"TSO                ",
	"TxCsumOffload      ",
	"RxCsumGood         ",
	"VLANextractions    ",
	"VLANinsertions     ",
	"GROpackets         ",
	"GROmerged          ",
};

static int get_sset_count(struct net_device *dev, int sset)
{
	switch (sset) {
	case ETH_SS_STATS:
		return ARRAY_SIZE(stats_strings);
	default:
		return -EOPNOTSUPP;
	}
}

#define T4_REGMAP_SIZE (160 * 1024)

static int get_regs_len(struct net_device *dev)
{
	return T4_REGMAP_SIZE;
}

static int get_eeprom_len(struct net_device *dev)
{
	return EEPROMSIZE;
}

static void get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info)
{
	struct adapter *adapter = netdev2adap(dev);

	strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver));
	strlcpy(info->version, DRV_VERSION, sizeof(info->version));
	strlcpy(info->bus_info, pci_name(adapter->pdev),
		sizeof(info->bus_info));

	if (adapter->params.fw_vers)
		snprintf(info->fw_version, sizeof(info->fw_version),
			"%u.%u.%u.%u, TP %u.%u.%u.%u",
			FW_HDR_FW_VER_MAJOR_GET(adapter->params.fw_vers),
			FW_HDR_FW_VER_MINOR_GET(adapter->params.fw_vers),
			FW_HDR_FW_VER_MICRO_GET(adapter->params.fw_vers),
			FW_HDR_FW_VER_BUILD_GET(adapter->params.fw_vers),
			FW_HDR_FW_VER_MAJOR_GET(adapter->params.tp_vers),
			FW_HDR_FW_VER_MINOR_GET(adapter->params.tp_vers),
			FW_HDR_FW_VER_MICRO_GET(adapter->params.tp_vers),
			FW_HDR_FW_VER_BUILD_GET(adapter->params.tp_vers));
}

static void get_strings(struct net_device *dev, u32 stringset, u8 *data)
{
	if (stringset == ETH_SS_STATS)
		memcpy(data, stats_strings, sizeof(stats_strings));
}

/*
 * port stats maintained per queue of the port.  They should be in the same
 * order as in stats_strings above.
 */
struct queue_port_stats {
	u64 tso;
	u64 tx_csum;
	u64 rx_csum;
	u64 vlan_ex;
	u64 vlan_ins;
	u64 gro_pkts;
	u64 gro_merged;
};

static void collect_sge_port_stats(const struct adapter *adap,
		const struct port_info *p, struct queue_port_stats *s)
{
	int i;
	const struct sge_eth_txq *tx = &adap->sge.ethtxq[p->first_qset];
	const struct sge_eth_rxq *rx = &adap->sge.ethrxq[p->first_qset];

	memset(s, 0, sizeof(*s));
	for (i = 0; i < p->nqsets; i++, rx++, tx++) {
		s->tso += tx->tso;
		s->tx_csum += tx->tx_cso;
		s->rx_csum += rx->stats.rx_cso;
		s->vlan_ex += rx->stats.vlan_ex;
		s->vlan_ins += tx->vlan_ins;
		s->gro_pkts += rx->stats.lro_pkts;
		s->gro_merged += rx->stats.lro_merged;
	}
}

static void get_stats(struct net_device *dev, struct ethtool_stats *stats,
		      u64 *data)
{
	struct port_info *pi = netdev_priv(dev);
	struct adapter *adapter = pi->adapter;

	t4_get_port_stats(adapter, pi->tx_chan, (struct port_stats *)data);

	data += sizeof(struct port_stats) / sizeof(u64);
	collect_sge_port_stats(adapter, pi, (struct queue_port_stats *)data);
}

/*
 * Return a version number to identify the type of adapter.  The scheme is:
 * - bits 0..9: chip version
 * - bits 10..15: chip revision
 * - bits 16..23: register dump version
 */
static inline unsigned int mk_adap_vers(const struct adapter *ap)
{
	return 4 | (ap->params.rev << 10) | (1 << 16);
}

static void reg_block_dump(struct adapter *ap, void *buf, unsigned int start,
			   unsigned int end)
{
	u32 *p = buf + start;

	for ( ; start <= end; start += sizeof(u32))
		*p++ = t4_read_reg(ap, start);
}

static void get_regs(struct net_device *dev, struct ethtool_regs *regs,
		     void *buf)
{
	static const unsigned int reg_ranges[] = {
		0x1008, 0x1108,
		0x1180, 0x11b4,
		0x11fc, 0x123c,
		0x1300, 0x173c,
		0x1800, 0x18fc,
		0x3000, 0x30d8,
		0x30e0, 0x5924,
		0x5960, 0x59d4,
		0x5a00, 0x5af8,
		0x6000, 0x6098,
		0x6100, 0x6150,
		0x6200, 0x6208,
		0x6240, 0x6248,
		0x6280, 0x6338,
		0x6370, 0x638c,
		0x6400, 0x643c,
		0x6500, 0x6524,
		0x6a00, 0x6a38,
		0x6a60, 0x6a78,
		0x6b00, 0x6b84,
		0x6bf0, 0x6c84,
		0x6cf0, 0x6d84,
		0x6df0, 0x6e84,
		0x6ef0, 0x6f84,
		0x6ff0, 0x7084,
		0x70f0, 0x7184,
		0x71f0, 0x7284,
		0x72f0, 0x7384,
		0x73f0, 0x7450,
		0x7500, 0x7530,
		0x7600, 0x761c,
		0x7680, 0x76cc,
		0x7700, 0x7798,
		0x77c0, 0x77fc,
		0x7900, 0x79fc,
		0x7b00, 0x7c38,
		0x7d00, 0x7efc,
		0x8dc0, 0x8e1c,
		0x8e30, 0x8e78,
		0x8ea0, 0x8f6c,
		0x8fc0, 0x9074,
		0x90fc, 0x90fc,
		0x9400, 0x9458,
		0x9600, 0x96bc,
		0x9800, 0x9808,
		0x9820, 0x983c,
		0x9850, 0x9864,
		0x9c00, 0x9c6c,
		0x9c80, 0x9cec,
		0x9d00, 0x9d6c,
		0x9d80, 0x9dec,
		0x9e00, 0x9e6c,
		0x9e80, 0x9eec,
		0x9f00, 0x9f6c,
		0x9f80, 0x9fec,
		0xd004, 0xd03c,
		0xdfc0, 0xdfe0,
		0xe000, 0xea7c,
		0xf000, 0x11190,
		0x19040, 0x1906c,
		0x19078, 0x19080,
		0x1908c, 0x19124,
		0x19150, 0x191b0,
		0x191d0, 0x191e8,
		0x19238, 0x1924c,
		0x193f8, 0x19474,
		0x19490, 0x194f8,
		0x19800, 0x19f30,
		0x1a000, 0x1a06c,
		0x1a0b0, 0x1a120,
		0x1a128, 0x1a138,
		0x1a190, 0x1a1c4,
		0x1a1fc, 0x1a1fc,
		0x1e040, 0x1e04c,
		0x1e284, 0x1e28c,
		0x1e2c0, 0x1e2c0,
		0x1e2e0, 0x1e2e0,
		0x1e300, 0x1e384,
		0x1e3c0, 0x1e3c8,
		0x1e440, 0x1e44c,
		0x1e684, 0x1e68c,
		0x1e6c0, 0x1e6c0,
		0x1e6e0, 0x1e6e0,
		0x1e700, 0x1e784,
		0x1e7c0, 0x1e7c8,
		0x1e840, 0x1e84c,
		0x1ea84, 0x1ea8c,
		0x1eac0, 0x1eac0,
		0x1eae0, 0x1eae0,
		0x1eb00, 0x1eb84,
		0x1ebc0, 0x1ebc8,
		0x1ec40, 0x1ec4c,
		0x1ee84, 0x1ee8c,
		0x1eec0, 0x1eec0,
		0x1eee0, 0x1eee0,
		0x1ef00, 0x1ef84,
		0x1efc0, 0x1efc8,
		0x1f040, 0x1f04c,
		0x1f284, 0x1f28c,
		0x1f2c0, 0x1f2c0,
		0x1f2e0, 0x1f2e0,
		0x1f300, 0x1f384,
		0x1f3c0, 0x1f3c8,
		0x1f440, 0x1f44c,
		0x1f684, 0x1f68c,
		0x1f6c0, 0x1f6c0,
		0x1f6e0, 0x1f6e0,
		0x1f700, 0x1f784,
		0x1f7c0, 0x1f7c8,
		0x1f840, 0x1f84c,
		0x1fa84, 0x1fa8c,
		0x1fac0, 0x1fac0,
		0x1fae0, 0x1fae0,
		0x1fb00, 0x1fb84,
		0x1fbc0, 0x1fbc8,
		0x1fc40, 0x1fc4c,
		0x1fe84, 0x1fe8c,
		0x1fec0, 0x1fec0,
		0x1fee0, 0x1fee0,
		0x1ff00, 0x1ff84,
		0x1ffc0, 0x1ffc8,
		0x20000, 0x2002c,
		0x20100, 0x2013c,
		0x20190, 0x201c8,
		0x20200, 0x20318,
		0x20400, 0x20528,
		0x20540, 0x20614,
		0x21000, 0x21040,
		0x2104c, 0x21060,
		0x210c0, 0x210ec,
		0x21200, 0x21268,
		0x21270, 0x21284,
		0x212fc, 0x21388,
		0x21400, 0x21404,
		0x21500, 0x21518,
		0x2152c, 0x2153c,
		0x21550, 0x21554,
		0x21600, 0x21600,
		0x21608, 0x21628,
		0x21630, 0x2163c,
		0x21700, 0x2171c,
		0x21780, 0x2178c,
		0x21800, 0x21c38,
		0x21c80, 0x21d7c,
		0x21e00, 0x21e04,
		0x22000, 0x2202c,
		0x22100, 0x2213c,
		0x22190, 0x221c8,
		0x22200, 0x22318,
		0x22400, 0x22528,
		0x22540, 0x22614,
		0x23000, 0x23040,
		0x2304c, 0x23060,
		0x230c0, 0x230ec,
		0x23200, 0x23268,
		0x23270, 0x23284,
		0x232fc, 0x23388,
		0x23400, 0x23404,
		0x23500, 0x23518,
		0x2352c, 0x2353c,
		0x23550, 0x23554,
		0x23600, 0x23600,
		0x23608, 0x23628,
		0x23630, 0x2363c,
		0x23700, 0x2371c,
		0x23780, 0x2378c,
		0x23800, 0x23c38,
		0x23c80, 0x23d7c,
		0x23e00, 0x23e04,
		0x24000, 0x2402c,
		0x24100, 0x2413c,
		0x24190, 0x241c8,
		0x24200, 0x24318,
		0x24400, 0x24528,
		0x24540, 0x24614,
		0x25000, 0x25040,
		0x2504c, 0x25060,
		0x250c0, 0x250ec,
		0x25200, 0x25268,
		0x25270, 0x25284,
		0x252fc, 0x25388,
		0x25400, 0x25404,
		0x25500, 0x25518,
		0x2552c, 0x2553c,
		0x25550, 0x25554,
		0x25600, 0x25600,
		0x25608, 0x25628,
		0x25630, 0x2563c,
		0x25700, 0x2571c,
		0x25780, 0x2578c,
		0x25800, 0x25c38,
		0x25c80, 0x25d7c,
		0x25e00, 0x25e04,
		0x26000, 0x2602c,
		0x26100, 0x2613c,
		0x26190, 0x261c8,
		0x26200, 0x26318,
		0x26400, 0x26528,
		0x26540, 0x26614,
		0x27000, 0x27040,
		0x2704c, 0x27060,
		0x270c0, 0x270ec,
		0x27200, 0x27268,
		0x27270, 0x27284,
		0x272fc, 0x27388,
		0x27400, 0x27404,
		0x27500, 0x27518,
		0x2752c, 0x2753c,
		0x27550, 0x27554,
		0x27600, 0x27600,
		0x27608, 0x27628,
		0x27630, 0x2763c,
		0x27700, 0x2771c,
		0x27780, 0x2778c,
		0x27800, 0x27c38,
		0x27c80, 0x27d7c,
		0x27e00, 0x27e04
	};

	int i;
	struct adapter *ap = netdev2adap(dev);

	regs->version = mk_adap_vers(ap);

	memset(buf, 0, T4_REGMAP_SIZE);
	for (i = 0; i < ARRAY_SIZE(reg_ranges); i += 2)
		reg_block_dump(ap, buf, reg_ranges[i], reg_ranges[i + 1]);
}

static int restart_autoneg(struct net_device *dev)
{
	struct port_info *p = netdev_priv(dev);

	if (!netif_running(dev))
		return -EAGAIN;
	if (p->link_cfg.autoneg != AUTONEG_ENABLE)
		return -EINVAL;
	t4_restart_aneg(p->adapter, p->adapter->fn, p->tx_chan);
	return 0;
}

static int identify_port(struct net_device *dev,
			 enum ethtool_phys_id_state state)
{
	unsigned int val;
	struct adapter *adap = netdev2adap(dev);

	if (state == ETHTOOL_ID_ACTIVE)
		val = 0xffff;
	else if (state == ETHTOOL_ID_INACTIVE)
		val = 0;
	else
		return -EINVAL;

	return t4_identify_port(adap, adap->fn, netdev2pinfo(dev)->viid, val);
}

static unsigned int from_fw_linkcaps(unsigned int type, unsigned int caps)
{
	unsigned int v = 0;

	if (type == FW_PORT_TYPE_BT_SGMII || type == FW_PORT_TYPE_BT_XFI ||
	    type == FW_PORT_TYPE_BT_XAUI) {
		v |= SUPPORTED_TP;
		if (caps & FW_PORT_CAP_SPEED_100M)
			v |= SUPPORTED_100baseT_Full;
		if (caps & FW_PORT_CAP_SPEED_1G)
			v |= SUPPORTED_1000baseT_Full;
		if (caps & FW_PORT_CAP_SPEED_10G)
			v |= SUPPORTED_10000baseT_Full;
	} else if (type == FW_PORT_TYPE_KX4 || type == FW_PORT_TYPE_KX) {
		v |= SUPPORTED_Backplane;
		if (caps & FW_PORT_CAP_SPEED_1G)
			v |= SUPPORTED_1000baseKX_Full;
		if (caps & FW_PORT_CAP_SPEED_10G)
			v |= SUPPORTED_10000baseKX4_Full;
	} else if (type == FW_PORT_TYPE_KR)
		v |= SUPPORTED_Backplane | SUPPORTED_10000baseKR_Full;
	else if (type == FW_PORT_TYPE_BP_AP)
		v |= SUPPORTED_Backplane | SUPPORTED_10000baseR_FEC |
		     SUPPORTED_10000baseKR_Full | SUPPORTED_1000baseKX_Full;
	else if (type == FW_PORT_TYPE_BP4_AP)
		v |= SUPPORTED_Backplane | SUPPORTED_10000baseR_FEC |
		     SUPPORTED_10000baseKR_Full | SUPPORTED_1000baseKX_Full |
		     SUPPORTED_10000baseKX4_Full;
	else if (type == FW_PORT_TYPE_FIBER_XFI ||
		 type == FW_PORT_TYPE_FIBER_XAUI || type == FW_PORT_TYPE_SFP)
		v |= SUPPORTED_FIBRE;

	if (caps & FW_PORT_CAP_ANEG)
		v |= SUPPORTED_Autoneg;
	return v;
}

static unsigned int to_fw_linkcaps(unsigned int caps)
{
	unsigned int v = 0;

	if (caps & ADVERTISED_100baseT_Full)
		v |= FW_PORT_CAP_SPEED_100M;
	if (caps & ADVERTISED_1000baseT_Full)
		v |= FW_PORT_CAP_SPEED_1G;
	if (caps & ADVERTISED_10000baseT_Full)
		v |= FW_PORT_CAP_SPEED_10G;
	return v;
}

static int get_settings(struct net_device *dev, struct ethtool_cmd *cmd)
{
	const struct port_info *p = netdev_priv(dev);

	if (p->port_type == FW_PORT_TYPE_BT_SGMII ||
	    p->port_type == FW_PORT_TYPE_BT_XFI ||
	    p->port_type == FW_PORT_TYPE_BT_XAUI)
		cmd->port = PORT_TP;
	else if (p->port_type == FW_PORT_TYPE_FIBER_XFI ||
		 p->port_type == FW_PORT_TYPE_FIBER_XAUI)
		cmd->port = PORT_FIBRE;
	else if (p->port_type == FW_PORT_TYPE_SFP) {
		if (p->mod_type == FW_PORT_MOD_TYPE_TWINAX_PASSIVE ||
		    p->mod_type == FW_PORT_MOD_TYPE_TWINAX_ACTIVE)
			cmd->port = PORT_DA;
		else
			cmd->port = PORT_FIBRE;
	} else
		cmd->port = PORT_OTHER;

	if (p->mdio_addr >= 0) {
		cmd->phy_address = p->mdio_addr;
		cmd->transceiver = XCVR_EXTERNAL;
		cmd->mdio_support = p->port_type == FW_PORT_TYPE_BT_SGMII ?
			MDIO_SUPPORTS_C22 : MDIO_SUPPORTS_C45;
	} else {
		cmd->phy_address = 0;  /* not really, but no better option */
		cmd->transceiver = XCVR_INTERNAL;
		cmd->mdio_support = 0;
	}

	cmd->supported = from_fw_linkcaps(p->port_type, p->link_cfg.supported);
	cmd->advertising = from_fw_linkcaps(p->port_type,
					    p->link_cfg.advertising);
	ethtool_cmd_speed_set(cmd,
			      netif_carrier_ok(dev) ? p->link_cfg.speed : 0);
	cmd->duplex = DUPLEX_FULL;
	cmd->autoneg = p->link_cfg.autoneg;
	cmd->maxtxpkt = 0;
	cmd->maxrxpkt = 0;
	return 0;
}

static unsigned int speed_to_caps(int speed)
{
	if (speed == SPEED_100)
		return FW_PORT_CAP_SPEED_100M;
	if (speed == SPEED_1000)
		return FW_PORT_CAP_SPEED_1G;
	if (speed == SPEED_10000)
		return FW_PORT_CAP_SPEED_10G;
	return 0;
}

static int set_settings(struct net_device *dev, struct ethtool_cmd *cmd)
{
	unsigned int cap;
	struct port_info *p = netdev_priv(dev);
	struct link_config *lc = &p->link_cfg;
	u32 speed = ethtool_cmd_speed(cmd);

	if (cmd->duplex != DUPLEX_FULL)     /* only full-duplex supported */
		return -EINVAL;

	if (!(lc->supported & FW_PORT_CAP_ANEG)) {
		/*
		 * PHY offers a single speed.  See if that's what's
		 * being requested.
		 */
		if (cmd->autoneg == AUTONEG_DISABLE &&
		    (lc->supported & speed_to_caps(speed)))
			return 0;
		return -EINVAL;
	}

	if (cmd->autoneg == AUTONEG_DISABLE) {
		cap = speed_to_caps(speed);

		if (!(lc->supported & cap) || (speed == SPEED_1000) ||
		    (speed == SPEED_10000))
			return -EINVAL;
		lc->requested_speed = cap;
		lc->advertising = 0;
	} else {
		cap = to_fw_linkcaps(cmd->advertising);
		if (!(lc->supported & cap))
			return -EINVAL;
		lc->requested_speed = 0;
		lc->advertising = cap | FW_PORT_CAP_ANEG;
	}
	lc->autoneg = cmd->autoneg;

	if (netif_running(dev))
		return t4_link_start(p->adapter, p->adapter->fn, p->tx_chan,
				     lc);
	return 0;
}

static void get_pauseparam(struct net_device *dev,
			   struct ethtool_pauseparam *epause)
{
	struct port_info *p = netdev_priv(dev);

	epause->autoneg = (p->link_cfg.requested_fc & PAUSE_AUTONEG) != 0;
	epause->rx_pause = (p->link_cfg.fc & PAUSE_RX) != 0;
	epause->tx_pause = (p->link_cfg.fc & PAUSE_TX) != 0;
}

static int set_pauseparam(struct net_device *dev,
			  struct ethtool_pauseparam *epause)
{
	struct port_info *p = netdev_priv(dev);
	struct link_config *lc = &p->link_cfg;

	if (epause->autoneg == AUTONEG_DISABLE)
		lc->requested_fc = 0;
	else if (lc->supported & FW_PORT_CAP_ANEG)
		lc->requested_fc = PAUSE_AUTONEG;
	else
		return -EINVAL;

	if (epause->rx_pause)
		lc->requested_fc |= PAUSE_RX;
	if (epause->tx_pause)
		lc->requested_fc |= PAUSE_TX;
	if (netif_running(dev))
		return t4_link_start(p->adapter, p->adapter->fn, p->tx_chan,
				     lc);
	return 0;
}

static void get_sge_param(struct net_device *dev, struct ethtool_ringparam *e)
{
	const struct port_info *pi = netdev_priv(dev);
	const struct sge *s = &pi->adapter->sge;

	e->rx_max_pending = MAX_RX_BUFFERS;
	e->rx_mini_max_pending = MAX_RSPQ_ENTRIES;
	e->rx_jumbo_max_pending = 0;
	e->tx_max_pending = MAX_TXQ_ENTRIES;

	e->rx_pending = s->ethrxq[pi->first_qset].fl.size - 8;
	e->rx_mini_pending = s->ethrxq[pi->first_qset].rspq.size;
	e->rx_jumbo_pending = 0;
	e->tx_pending = s->ethtxq[pi->first_qset].q.size;
}

static int set_sge_param(struct net_device *dev, struct ethtool_ringparam *e)
{
	int i;
	const struct port_info *pi = netdev_priv(dev);
	struct adapter *adapter = pi->adapter;
	struct sge *s = &adapter->sge;

	if (e->rx_pending > MAX_RX_BUFFERS || e->rx_jumbo_pending ||
	    e->tx_pending > MAX_TXQ_ENTRIES ||
	    e->rx_mini_pending > MAX_RSPQ_ENTRIES ||
	    e->rx_mini_pending < MIN_RSPQ_ENTRIES ||
	    e->rx_pending < MIN_FL_ENTRIES || e->tx_pending < MIN_TXQ_ENTRIES)
		return -EINVAL;

	if (adapter->flags & FULL_INIT_DONE)
		return -EBUSY;

	for (i = 0; i < pi->nqsets; ++i) {
		s->ethtxq[pi->first_qset + i].q.size = e->tx_pending;
		s->ethrxq[pi->first_qset + i].fl.size = e->rx_pending + 8;
		s->ethrxq[pi->first_qset + i].rspq.size = e->rx_mini_pending;
	}
	return 0;
}

static int closest_timer(const struct sge *s, int time)
{
	int i, delta, match = 0, min_delta = INT_MAX;

	for (i = 0; i < ARRAY_SIZE(s->timer_val); i++) {
		delta = time - s->timer_val[i];
		if (delta < 0)
			delta = -delta;
		if (delta < min_delta) {
			min_delta = delta;
			match = i;
		}
	}
	return match;
}

static int closest_thres(const struct sge *s, int thres)
{
	int i, delta, match = 0, min_delta = INT_MAX;

	for (i = 0; i < ARRAY_SIZE(s->counter_val); i++) {
		delta = thres - s->counter_val[i];
		if (delta < 0)
			delta = -delta;
		if (delta < min_delta) {
			min_delta = delta;
			match = i;
		}
	}
	return match;
}

/*
 * Return a queue's interrupt hold-off time in us.  0 means no timer.
 */
static unsigned int qtimer_val(const struct adapter *adap,
			       const struct sge_rspq *q)
{
	unsigned int idx = q->intr_params >> 1;

	return idx < SGE_NTIMERS ? adap->sge.timer_val[idx] : 0;
}

/**
 *	set_rxq_intr_params - set a queue's interrupt holdoff parameters
 *	@adap: the adapter
 *	@q: the Rx queue
 *	@us: the hold-off time in us, or 0 to disable timer
 *	@cnt: the hold-off packet count, or 0 to disable counter
 *
 *	Sets an Rx queue's interrupt hold-off time and packet count.  At least
 *	one of the two needs to be enabled for the queue to generate interrupts.
 */
static int set_rxq_intr_params(struct adapter *adap, struct sge_rspq *q,
			       unsigned int us, unsigned int cnt)
{
	if ((us | cnt) == 0)
		cnt = 1;

	if (cnt) {
		int err;
		u32 v, new_idx;

		new_idx = closest_thres(&adap->sge, cnt);
		if (q->desc && q->pktcnt_idx != new_idx) {
			/* the queue has already been created, update it */
			v = FW_PARAMS_MNEM(FW_PARAMS_MNEM_DMAQ) |
			    FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DMAQ_IQ_INTCNTTHRESH) |
			    FW_PARAMS_PARAM_YZ(q->cntxt_id);
			err = t4_set_params(adap, adap->fn, adap->fn, 0, 1, &v,
					    &new_idx);
			if (err)
				return err;
		}
		q->pktcnt_idx = new_idx;
	}

	us = us == 0 ? 6 : closest_timer(&adap->sge, us);
	q->intr_params = QINTR_TIMER_IDX(us) | (cnt > 0 ? QINTR_CNT_EN : 0);
	return 0;
}

static int set_coalesce(struct net_device *dev, struct ethtool_coalesce *c)
{
	const struct port_info *pi = netdev_priv(dev);
	struct adapter *adap = pi->adapter;

	return set_rxq_intr_params(adap, &adap->sge.ethrxq[pi->first_qset].rspq,
			c->rx_coalesce_usecs, c->rx_max_coalesced_frames);
}

static int get_coalesce(struct net_device *dev, struct ethtool_coalesce *c)
{
	const struct port_info *pi = netdev_priv(dev);
	const struct adapter *adap = pi->adapter;
	const struct sge_rspq *rq = &adap->sge.ethrxq[pi->first_qset].rspq;

	c->rx_coalesce_usecs = qtimer_val(adap, rq);
	c->rx_max_coalesced_frames = (rq->intr_params & QINTR_CNT_EN) ?
		adap->sge.counter_val[rq->pktcnt_idx] : 0;
	return 0;
}

/**
 *	eeprom_ptov - translate a physical EEPROM address to virtual
 *	@phys_addr: the physical EEPROM address
 *	@fn: the PCI function number
 *	@sz: size of function-specific area
 *
 *	Translate a physical EEPROM address to virtual.  The first 1K is
 *	accessed through virtual addresses starting at 31K, the rest is
 *	accessed through virtual addresses starting at 0.
 *
 *	The mapping is as follows:
 *	[0..1K) -> [31K..32K)
 *	[1K..1K+A) -> [31K-A..31K)
 *	[1K+A..ES) -> [0..ES-A-1K)
 *
 *	where A = @fn * @sz, and ES = EEPROM size.
 */
static int eeprom_ptov(unsigned int phys_addr, unsigned int fn, unsigned int sz)
{
	fn *= sz;
	if (phys_addr < 1024)
		return phys_addr + (31 << 10);
	if (phys_addr < 1024 + fn)
		return 31744 - fn + phys_addr - 1024;
	if (phys_addr < EEPROMSIZE)
		return phys_addr - 1024 - fn;
	return -EINVAL;
}

/*
 * The next two routines implement eeprom read/write from physical addresses.
 */
static int eeprom_rd_phys(struct adapter *adap, unsigned int phys_addr, u32 *v)
{
	int vaddr = eeprom_ptov(phys_addr, adap->fn, EEPROMPFSIZE);

	if (vaddr >= 0)
		vaddr = pci_read_vpd(adap->pdev, vaddr, sizeof(u32), v);
	return vaddr < 0 ? vaddr : 0;
}

static int eeprom_wr_phys(struct adapter *adap, unsigned int phys_addr, u32 v)
{
	int vaddr = eeprom_ptov(phys_addr, adap->fn, EEPROMPFSIZE);

	if (vaddr >= 0)
		vaddr = pci_write_vpd(adap->pdev, vaddr, sizeof(u32), &v);
	return vaddr < 0 ? vaddr : 0;
}

#define EEPROM_MAGIC 0x38E2F10C

static int get_eeprom(struct net_device *dev, struct ethtool_eeprom *e,
		      u8 *data)
{
	int i, err = 0;
	struct adapter *adapter = netdev2adap(dev);

	u8 *buf = kmalloc(EEPROMSIZE, GFP_KERNEL);
	if (!buf)
		return -ENOMEM;

	e->magic = EEPROM_MAGIC;
	for (i = e->offset & ~3; !err && i < e->offset + e->len; i += 4)
		err = eeprom_rd_phys(adapter, i, (u32 *)&buf[i]);

	if (!err)
		memcpy(data, buf + e->offset, e->len);
	kfree(buf);
	return err;
}

static int set_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom,
		      u8 *data)
{
	u8 *buf;
	int err = 0;
	u32 aligned_offset, aligned_len, *p;
	struct adapter *adapter = netdev2adap(dev);

	if (eeprom->magic != EEPROM_MAGIC)
		return -EINVAL;

	aligned_offset = eeprom->offset & ~3;
	aligned_len = (eeprom->len + (eeprom->offset & 3) + 3) & ~3;

	if (adapter->fn > 0) {
		u32 start = 1024 + adapter->fn * EEPROMPFSIZE;

		if (aligned_offset < start ||
		    aligned_offset + aligned_len > start + EEPROMPFSIZE)
			return -EPERM;
	}

	if (aligned_offset != eeprom->offset || aligned_len != eeprom->len) {
		/*
		 * RMW possibly needed for first or last words.
		 */
		buf = kmalloc(aligned_len, GFP_KERNEL);
		if (!buf)
			return -ENOMEM;
		err = eeprom_rd_phys(adapter, aligned_offset, (u32 *)buf);
		if (!err && aligned_len > 4)
			err = eeprom_rd_phys(adapter,
					     aligned_offset + aligned_len - 4,
					     (u32 *)&buf[aligned_len - 4]);
		if (err)
			goto out;
		memcpy(buf + (eeprom->offset & 3), data, eeprom->len);
	} else
		buf = data;

	err = t4_seeprom_wp(adapter, false);
	if (err)
		goto out;

	for (p = (u32 *)buf; !err && aligned_len; aligned_len -= 4, p++) {
		err = eeprom_wr_phys(adapter, aligned_offset, *p);
		aligned_offset += 4;
	}

	if (!err)
		err = t4_seeprom_wp(adapter, true);
out:
	if (buf != data)
		kfree(buf);
	return err;
}

static int set_flash(struct net_device *netdev, struct ethtool_flash *ef)
{
	int ret;
	const struct firmware *fw;
	struct adapter *adap = netdev2adap(netdev);

	ef->data[sizeof(ef->data) - 1] = '\0';
	ret = request_firmware(&fw, ef->data, adap->pdev_dev);
	if (ret < 0)
		return ret;

	ret = t4_load_fw(adap, fw->data, fw->size);
	release_firmware(fw);
	if (!ret)
		dev_info(adap->pdev_dev, "loaded firmware %s\n", ef->data);
	return ret;
}

#define WOL_SUPPORTED (WAKE_BCAST | WAKE_MAGIC)
#define BCAST_CRC 0xa0ccc1a6

static void get_wol(struct net_device *dev, struct ethtool_wolinfo *wol)
{
	wol->supported = WAKE_BCAST | WAKE_MAGIC;
	wol->wolopts = netdev2adap(dev)->wol;
	memset(&wol->sopass, 0, sizeof(wol->sopass));
}

static int set_wol(struct net_device *dev, struct ethtool_wolinfo *wol)
{
	int err = 0;
	struct port_info *pi = netdev_priv(dev);

	if (wol->wolopts & ~WOL_SUPPORTED)
		return -EINVAL;
	t4_wol_magic_enable(pi->adapter, pi->tx_chan,
			    (wol->wolopts & WAKE_MAGIC) ? dev->dev_addr : NULL);
	if (wol->wolopts & WAKE_BCAST) {
		err = t4_wol_pat_enable(pi->adapter, pi->tx_chan, 0xfe, ~0ULL,
					~0ULL, 0, false);
		if (!err)
			err = t4_wol_pat_enable(pi->adapter, pi->tx_chan, 1,
						~6ULL, ~0ULL, BCAST_CRC, true);
	} else
		t4_wol_pat_enable(pi->adapter, pi->tx_chan, 0, 0, 0, 0, false);
	return err;
}

static int cxgb_set_features(struct net_device *dev, netdev_features_t features)
{
	const struct port_info *pi = netdev_priv(dev);
	netdev_features_t changed = dev->features ^ features;
	int err;

	if (!(changed & NETIF_F_HW_VLAN_RX))
		return 0;

	err = t4_set_rxmode(pi->adapter, pi->adapter->fn, pi->viid, -1,
			    -1, -1, -1,
			    !!(features & NETIF_F_HW_VLAN_RX), true);
	if (unlikely(err))
		dev->features = features ^ NETIF_F_HW_VLAN_RX;
	return err;
}

static u32 get_rss_table_size(struct net_device *dev)
{
	const struct port_info *pi = netdev_priv(dev);

	return pi->rss_size;
}

static int get_rss_table(struct net_device *dev, u32 *p)
{
	const struct port_info *pi = netdev_priv(dev);
	unsigned int n = pi->rss_size;

	while (n--)
		p[n] = pi->rss[n];
	return 0;
}

static int set_rss_table(struct net_device *dev, const u32 *p)
{
	unsigned int i;
	struct port_info *pi = netdev_priv(dev);

	for (i = 0; i < pi->rss_size; i++)
		pi->rss[i] = p[i];
	if (pi->adapter->flags & FULL_INIT_DONE)
		return write_rss(pi, pi->rss);
	return 0;
}

static int get_rxnfc(struct net_device *dev, struct ethtool_rxnfc *info,
		     u32 *rules)
{
	const struct port_info *pi = netdev_priv(dev);

	switch (info->cmd) {
	case ETHTOOL_GRXFH: {
		unsigned int v = pi->rss_mode;

		info->data = 0;
		switch (info->flow_type) {
		case TCP_V4_FLOW:
			if (v & FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN)
				info->data = RXH_IP_SRC | RXH_IP_DST |
					     RXH_L4_B_0_1 | RXH_L4_B_2_3;
			else if (v & FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN)
				info->data = RXH_IP_SRC | RXH_IP_DST;
			break;
		case UDP_V4_FLOW:
			if ((v & FW_RSS_VI_CONFIG_CMD_IP4FOURTUPEN) &&
			    (v & FW_RSS_VI_CONFIG_CMD_UDPEN))
				info->data = RXH_IP_SRC | RXH_IP_DST |
					     RXH_L4_B_0_1 | RXH_L4_B_2_3;
			else if (v & FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN)
				info->data = RXH_IP_SRC | RXH_IP_DST;
			break;
		case SCTP_V4_FLOW:
		case AH_ESP_V4_FLOW:
		case IPV4_FLOW:
			if (v & FW_RSS_VI_CONFIG_CMD_IP4TWOTUPEN)
				info->data = RXH_IP_SRC | RXH_IP_DST;
			break;
		case TCP_V6_FLOW:
			if (v & FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN)
				info->data = RXH_IP_SRC | RXH_IP_DST |
					     RXH_L4_B_0_1 | RXH_L4_B_2_3;
			else if (v & FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN)
				info->data = RXH_IP_SRC | RXH_IP_DST;
			break;
		case UDP_V6_FLOW:
			if ((v & FW_RSS_VI_CONFIG_CMD_IP6FOURTUPEN) &&
			    (v & FW_RSS_VI_CONFIG_CMD_UDPEN))
				info->data = RXH_IP_SRC | RXH_IP_DST |
					     RXH_L4_B_0_1 | RXH_L4_B_2_3;
			else if (v & FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN)
				info->data = RXH_IP_SRC | RXH_IP_DST;
			break;
		case SCTP_V6_FLOW:
		case AH_ESP_V6_FLOW:
		case IPV6_FLOW:
			if (v & FW_RSS_VI_CONFIG_CMD_IP6TWOTUPEN)
				info->data = RXH_IP_SRC | RXH_IP_DST;
			break;
		}
		return 0;
	}
	case ETHTOOL_GRXRINGS:
		info->data = pi->nqsets;
		return 0;
	}
	return -EOPNOTSUPP;
}

static const struct ethtool_ops cxgb_ethtool_ops = {
	.get_settings      = get_settings,
	.set_settings      = set_settings,
	.get_drvinfo       = get_drvinfo,
	.get_msglevel      = get_msglevel,
	.set_msglevel      = set_msglevel,
	.get_ringparam     = get_sge_param,
	.set_ringparam     = set_sge_param,
	.get_coalesce      = get_coalesce,
	.set_coalesce      = set_coalesce,
	.get_eeprom_len    = get_eeprom_len,
	.get_eeprom        = get_eeprom,
	.set_eeprom        = set_eeprom,
	.get_pauseparam    = get_pauseparam,
	.set_pauseparam    = set_pauseparam,
	.get_link          = ethtool_op_get_link,
	.get_strings       = get_strings,
	.set_phys_id       = identify_port,
	.nway_reset        = restart_autoneg,
	.get_sset_count    = get_sset_count,
	.get_ethtool_stats = get_stats,
	.get_regs_len      = get_regs_len,
	.get_regs          = get_regs,
	.get_wol           = get_wol,
	.set_wol           = set_wol,
	.get_rxnfc         = get_rxnfc,
	.get_rxfh_indir_size = get_rss_table_size,
	.get_rxfh_indir    = get_rss_table,
	.set_rxfh_indir    = set_rss_table,
	.flash_device      = set_flash,
};

/*
 * debugfs support
 */
static ssize_t mem_read(struct file *file, char __user *buf, size_t count,
			loff_t *ppos)
{
	loff_t pos = *ppos;
	loff_t avail = file->f_path.dentry->d_inode->i_size;
	unsigned int mem = (uintptr_t)file->private_data & 3;
	struct adapter *adap = file->private_data - mem;

	if (pos < 0)
		return -EINVAL;
	if (pos >= avail)
		return 0;
	if (count > avail - pos)
		count = avail - pos;

	while (count) {
		size_t len;
		int ret, ofst;
		__be32 data[16];

		if (mem == MEM_MC)
			ret = t4_mc_read(adap, pos, data, NULL);
		else
			ret = t4_edc_read(adap, mem, pos, data, NULL);
		if (ret)
			return ret;

		ofst = pos % sizeof(data);
		len = min(count, sizeof(data) - ofst);
		if (copy_to_user(buf, (u8 *)data + ofst, len))
			return -EFAULT;

		buf += len;
		pos += len;
		count -= len;
	}
	count = pos - *ppos;
	*ppos = pos;
	return count;
}

static const struct file_operations mem_debugfs_fops = {
	.owner   = THIS_MODULE,
	.open    = simple_open,
	.read    = mem_read,
	.llseek  = default_llseek,
};

static void add_debugfs_mem(struct adapter *adap, const char *name,
			    unsigned int idx, unsigned int size_mb)
{
	struct dentry *de;

	de = debugfs_create_file(name, S_IRUSR, adap->debugfs_root,
				 (void *)adap + idx, &mem_debugfs_fops);
	if (de && de->d_inode)
		de->d_inode->i_size = size_mb << 20;
}

static int setup_debugfs(struct adapter *adap)
{
	int i;

	if (IS_ERR_OR_NULL(adap->debugfs_root))
		return -1;

	i = t4_read_reg(adap, MA_TARGET_MEM_ENABLE);
	if (i & EDRAM0_ENABLE)
		add_debugfs_mem(adap, "edc0", MEM_EDC0, 5);
	if (i & EDRAM1_ENABLE)
		add_debugfs_mem(adap, "edc1", MEM_EDC1, 5);
	if (i & EXT_MEM_ENABLE)
		add_debugfs_mem(adap, "mc", MEM_MC,
			EXT_MEM_SIZE_GET(t4_read_reg(adap, MA_EXT_MEMORY_BAR)));
	if (adap->l2t)
		debugfs_create_file("l2t", S_IRUSR, adap->debugfs_root, adap,
				    &t4_l2t_fops);
	return 0;
}

/*
 * upper-layer driver support
 */

/*
 * Allocate an active-open TID and set it to the supplied value.
 */
int cxgb4_alloc_atid(struct tid_info *t, void *data)
{
	int atid = -1;

	spin_lock_bh(&t->atid_lock);
	if (t->afree) {
		union aopen_entry *p = t->afree;

		atid = (p - t->atid_tab) + t->atid_base;
		t->afree = p->next;
		p->data = data;
		t->atids_in_use++;
	}
	spin_unlock_bh(&t->atid_lock);
	return atid;
}
EXPORT_SYMBOL(cxgb4_alloc_atid);

/*
 * Release an active-open TID.
 */
void cxgb4_free_atid(struct tid_info *t, unsigned int atid)
{
	union aopen_entry *p = &t->atid_tab[atid - t->atid_base];

	spin_lock_bh(&t->atid_lock);
	p->next = t->afree;
	t->afree = p;
	t->atids_in_use--;
	spin_unlock_bh(&t->atid_lock);
}
EXPORT_SYMBOL(cxgb4_free_atid);

/*
 * Allocate a server TID and set it to the supplied value.
 */
int cxgb4_alloc_stid(struct tid_info *t, int family, void *data)
{
	int stid;

	spin_lock_bh(&t->stid_lock);
	if (family == PF_INET) {
		stid = find_first_zero_bit(t->stid_bmap, t->nstids);
		if (stid < t->nstids)
			__set_bit(stid, t->stid_bmap);
		else
			stid = -1;
	} else {
		stid = bitmap_find_free_region(t->stid_bmap, t->nstids, 2);
		if (stid < 0)
			stid = -1;
	}
	if (stid >= 0) {
		t->stid_tab[stid].data = data;
		stid += t->stid_base;
		t->stids_in_use++;
	}
	spin_unlock_bh(&t->stid_lock);
	return stid;
}
EXPORT_SYMBOL(cxgb4_alloc_stid);

/*
 * Release a server TID.
 */
void cxgb4_free_stid(struct tid_info *t, unsigned int stid, int family)
{
	stid -= t->stid_base;
	spin_lock_bh(&t->stid_lock);
	if (family == PF_INET)
		__clear_bit(stid, t->stid_bmap);
	else
		bitmap_release_region(t->stid_bmap, stid, 2);
	t->stid_tab[stid].data = NULL;
	t->stids_in_use--;
	spin_unlock_bh(&t->stid_lock);
}
EXPORT_SYMBOL(cxgb4_free_stid);

/*
 * Populate a TID_RELEASE WR.  Caller must properly size the skb.
 */
static void mk_tid_release(struct sk_buff *skb, unsigned int chan,
			   unsigned int tid)
{
	struct cpl_tid_release *req;

	set_wr_txq(skb, CPL_PRIORITY_SETUP, chan);
	req = (struct cpl_tid_release *)__skb_put(skb, sizeof(*req));
	INIT_TP_WR(req, tid);
	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_TID_RELEASE, tid));
}

/*
 * Queue a TID release request and if necessary schedule a work queue to
 * process it.
 */
static void cxgb4_queue_tid_release(struct tid_info *t, unsigned int chan,
				    unsigned int tid)
{
	void **p = &t->tid_tab[tid];
	struct adapter *adap = container_of(t, struct adapter, tids);

	spin_lock_bh(&adap->tid_release_lock);
	*p = adap->tid_release_head;
	/* Low 2 bits encode the Tx channel number */
	adap->tid_release_head = (void **)((uintptr_t)p | chan);
	if (!adap->tid_release_task_busy) {
		adap->tid_release_task_busy = true;
		queue_work(workq, &adap->tid_release_task);
	}
	spin_unlock_bh(&adap->tid_release_lock);
}

/*
 * Process the list of pending TID release requests.
 */
static void process_tid_release_list(struct work_struct *work)
{
	struct sk_buff *skb;
	struct adapter *adap;

	adap = container_of(work, struct adapter, tid_release_task);

	spin_lock_bh(&adap->tid_release_lock);
	while (adap->tid_release_head) {
		void **p = adap->tid_release_head;
		unsigned int chan = (uintptr_t)p & 3;
		p = (void *)p - chan;

		adap->tid_release_head = *p;
		*p = NULL;
		spin_unlock_bh(&adap->tid_release_lock);

		while (!(skb = alloc_skb(sizeof(struct cpl_tid_release),
					 GFP_KERNEL)))
			schedule_timeout_uninterruptible(1);

		mk_tid_release(skb, chan, p - adap->tids.tid_tab);
		t4_ofld_send(adap, skb);
		spin_lock_bh(&adap->tid_release_lock);
	}
	adap->tid_release_task_busy = false;
	spin_unlock_bh(&adap->tid_release_lock);
}

/*
 * Release a TID and inform HW.  If we are unable to allocate the release
 * message we defer to a work queue.
 */
void cxgb4_remove_tid(struct tid_info *t, unsigned int chan, unsigned int tid)
{
	void *old;
	struct sk_buff *skb;
	struct adapter *adap = container_of(t, struct adapter, tids);

	old = t->tid_tab[tid];
	skb = alloc_skb(sizeof(struct cpl_tid_release), GFP_ATOMIC);
	if (likely(skb)) {
		t->tid_tab[tid] = NULL;
		mk_tid_release(skb, chan, tid);
		t4_ofld_send(adap, skb);
	} else
		cxgb4_queue_tid_release(t, chan, tid);
	if (old)
		atomic_dec(&t->tids_in_use);
}
EXPORT_SYMBOL(cxgb4_remove_tid);

/*
 * Allocate and initialize the TID tables.  Returns 0 on success.
 */
static int tid_init(struct tid_info *t)
{
	size_t size;
	unsigned int stid_bmap_size;
	unsigned int natids = t->natids;

	stid_bmap_size = BITS_TO_LONGS(t->nstids);
	size = t->ntids * sizeof(*t->tid_tab) +
	       natids * sizeof(*t->atid_tab) +
	       t->nstids * sizeof(*t->stid_tab) +
	       stid_bmap_size * sizeof(long) +
	       t->nftids * sizeof(*t->ftid_tab);

	t->tid_tab = t4_alloc_mem(size);
	if (!t->tid_tab)
		return -ENOMEM;

	t->atid_tab = (union aopen_entry *)&t->tid_tab[t->ntids];
	t->stid_tab = (struct serv_entry *)&t->atid_tab[natids];
	t->stid_bmap = (unsigned long *)&t->stid_tab[t->nstids];
	t->ftid_tab = (struct filter_entry *)&t->stid_bmap[stid_bmap_size];
	spin_lock_init(&t->stid_lock);
	spin_lock_init(&t->atid_lock);

	t->stids_in_use = 0;
	t->afree = NULL;
	t->atids_in_use = 0;
	atomic_set(&t->tids_in_use, 0);

	/* Setup the free list for atid_tab and clear the stid bitmap. */
	if (natids) {
		while (--natids)
			t->atid_tab[natids - 1].next = &t->atid_tab[natids];
		t->afree = t->atid_tab;
	}
	bitmap_zero(t->stid_bmap, t->nstids);
	return 0;
}

/**
 *	cxgb4_create_server - create an IP server
 *	@dev: the device
 *	@stid: the server TID
 *	@sip: local IP address to bind server to
 *	@sport: the server's TCP port
 *	@queue: queue to direct messages from this server to
 *
 *	Create an IP server for the given port and address.
 *	Returns <0 on error and one of the %NET_XMIT_* values on success.
 */
int cxgb4_create_server(const struct net_device *dev, unsigned int stid,
			__be32 sip, __be16 sport, unsigned int queue)
{
	unsigned int chan;
	struct sk_buff *skb;
	struct adapter *adap;
	struct cpl_pass_open_req *req;

	skb = alloc_skb(sizeof(*req), GFP_KERNEL);
	if (!skb)
		return -ENOMEM;

	adap = netdev2adap(dev);
	req = (struct cpl_pass_open_req *)__skb_put(skb, sizeof(*req));
	INIT_TP_WR(req, 0);
	OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_PASS_OPEN_REQ, stid));
	req->local_port = sport;
	req->peer_port = htons(0);
	req->local_ip = sip;
	req->peer_ip = htonl(0);
	chan = rxq_to_chan(&adap->sge, queue);
	req->opt0 = cpu_to_be64(TX_CHAN(chan));
	req->opt1 = cpu_to_be64(CONN_POLICY_ASK |
				SYN_RSS_ENABLE | SYN_RSS_QUEUE(queue));
	return t4_mgmt_tx(adap, skb);
}
EXPORT_SYMBOL(cxgb4_create_server);

/**
 *	cxgb4_best_mtu - find the entry in the MTU table closest to an MTU
 *	@mtus: the HW MTU table
 *	@mtu: the target MTU
 *	@idx: index of selected entry in the MTU table
 *
 *	Returns the index and the value in the HW MTU table that is closest to
 *	but does not exceed @mtu, unless @mtu is smaller than any value in the
 *	table, in which case that smallest available value is selected.
 */
unsigned int cxgb4_best_mtu(const unsigned short *mtus, unsigned short mtu,
			    unsigned int *idx)
{
	unsigned int i = 0;

	while (i < NMTUS - 1 && mtus[i + 1] <= mtu)
		++i;
	if (idx)
		*idx = i;
	return mtus[i];
}
EXPORT_SYMBOL(cxgb4_best_mtu);

/**
 *	cxgb4_port_chan - get the HW channel of a port
 *	@dev: the net device for the port
 *
 *	Return the HW Tx channel of the given port.
 */
unsigned int cxgb4_port_chan(const struct net_device *dev)
{
	return netdev2pinfo(dev)->tx_chan;
}
EXPORT_SYMBOL(cxgb4_port_chan);

unsigned int cxgb4_dbfifo_count(const struct net_device *dev, int lpfifo)
{
	struct adapter *adap = netdev2adap(dev);
	u32 v;

	v = t4_read_reg(adap, A_SGE_DBFIFO_STATUS);
	return lpfifo ? G_LP_COUNT(v) : G_HP_COUNT(v);
}
EXPORT_SYMBOL(cxgb4_dbfifo_count);

/**
 *	cxgb4_port_viid - get the VI id of a port
 *	@dev: the net device for the port
 *
 *	Return the VI id of the given port.
 */
unsigned int cxgb4_port_viid(const struct net_device *dev)
{
	return netdev2pinfo(dev)->viid;
}
EXPORT_SYMBOL(cxgb4_port_viid);

/**
 *	cxgb4_port_idx - get the index of a port
 *	@dev: the net device for the port
 *
 *	Return the index of the given port.
 */
unsigned int cxgb4_port_idx(const struct net_device *dev)
{
	return netdev2pinfo(dev)->port_id;
}
EXPORT_SYMBOL(cxgb4_port_idx);

void cxgb4_get_tcp_stats(struct pci_dev *pdev, struct tp_tcp_stats *v4,
			 struct tp_tcp_stats *v6)
{
	struct adapter *adap = pci_get_drvdata(pdev);

	spin_lock(&adap->stats_lock);
	t4_tp_get_tcp_stats(adap, v4, v6);
	spin_unlock(&adap->stats_lock);
}
EXPORT_SYMBOL(cxgb4_get_tcp_stats);

void cxgb4_iscsi_init(struct net_device *dev, unsigned int tag_mask,
		      const unsigned int *pgsz_order)
{
	struct adapter *adap = netdev2adap(dev);

	t4_write_reg(adap, ULP_RX_ISCSI_TAGMASK, tag_mask);
	t4_write_reg(adap, ULP_RX_ISCSI_PSZ, HPZ0(pgsz_order[0]) |
		     HPZ1(pgsz_order[1]) | HPZ2(pgsz_order[2]) |
		     HPZ3(pgsz_order[3]));
}
EXPORT_SYMBOL(cxgb4_iscsi_init);

int cxgb4_flush_eq_cache(struct net_device *dev)
{
	struct adapter *adap = netdev2adap(dev);
	int ret;

	ret = t4_fwaddrspace_write(adap, adap->mbox,
				   0xe1000000 + A_SGE_CTXT_CMD, 0x20000000);
	return ret;
}
EXPORT_SYMBOL(cxgb4_flush_eq_cache);

static int read_eq_indices(struct adapter *adap, u16 qid, u16 *pidx, u16 *cidx)
{
	u32 addr = t4_read_reg(adap, A_SGE_DBQ_CTXT_BADDR) + 24 * qid + 8;
	__be64 indices;
	int ret;

	ret = t4_mem_win_read_len(adap, addr, (__be32 *)&indices, 8);
	if (!ret) {
		*cidx = (be64_to_cpu(indices) >> 25) & 0xffff;
		*pidx = (be64_to_cpu(indices) >> 9) & 0xffff;
	}
	return ret;
}

int cxgb4_sync_txq_pidx(struct net_device *dev, u16 qid, u16 pidx,
			u16 size)
{
	struct adapter *adap = netdev2adap(dev);
	u16 hw_pidx, hw_cidx;
	int ret;

	ret = read_eq_indices(adap, qid, &hw_pidx, &hw_cidx);
	if (ret)
		goto out;

	if (pidx != hw_pidx) {
		u16 delta;

		if (pidx >= hw_pidx)
			delta = pidx - hw_pidx;
		else
			delta = size - hw_pidx + pidx;
		wmb();
		t4_write_reg(adap, MYPF_REG(SGE_PF_KDOORBELL),
			     QID(qid) | PIDX(delta));
	}
out:
	return ret;
}
EXPORT_SYMBOL(cxgb4_sync_txq_pidx);

static struct pci_driver cxgb4_driver;

static void check_neigh_update(struct neighbour *neigh)
{
	const struct device *parent;
	const struct net_device *netdev = neigh->dev;

	if (netdev->priv_flags & IFF_802_1Q_VLAN)
		netdev = vlan_dev_real_dev(netdev);
	parent = netdev->dev.parent;
	if (parent && parent->driver == &cxgb4_driver.driver)
		t4_l2t_update(dev_get_drvdata(parent), neigh);
}

static int netevent_cb(struct notifier_block *nb, unsigned long event,
		       void *data)
{
	switch (event) {
	case NETEVENT_NEIGH_UPDATE:
		check_neigh_update(data);
		break;
	case NETEVENT_REDIRECT:
	default:
		break;
	}
	return 0;
}

static bool netevent_registered;
static struct notifier_block cxgb4_netevent_nb = {
	.notifier_call = netevent_cb
};

static void drain_db_fifo(struct adapter *adap, int usecs)
{
	u32 v;

	do {
		set_current_state(TASK_UNINTERRUPTIBLE);
		schedule_timeout(usecs_to_jiffies(usecs));
		v = t4_read_reg(adap, A_SGE_DBFIFO_STATUS);
		if (G_LP_COUNT(v) == 0 && G_HP_COUNT(v) == 0)
			break;
	} while (1);
}

static void disable_txq_db(struct sge_txq *q)
{
	spin_lock_irq(&q->db_lock);
	q->db_disabled = 1;
	spin_unlock_irq(&q->db_lock);
}

static void enable_txq_db(struct sge_txq *q)
{
	spin_lock_irq(&q->db_lock);
	q->db_disabled = 0;
	spin_unlock_irq(&q->db_lock);
}

static void disable_dbs(struct adapter *adap)
{
	int i;

	for_each_ethrxq(&adap->sge, i)
		disable_txq_db(&adap->sge.ethtxq[i].q);
	for_each_ofldrxq(&adap->sge, i)
		disable_txq_db(&adap->sge.ofldtxq[i].q);
	for_each_port(adap, i)
		disable_txq_db(&adap->sge.ctrlq[i].q);
}

static void enable_dbs(struct adapter *adap)
{
	int i;

	for_each_ethrxq(&adap->sge, i)
		enable_txq_db(&adap->sge.ethtxq[i].q);
	for_each_ofldrxq(&adap->sge, i)
		enable_txq_db(&adap->sge.ofldtxq[i].q);
	for_each_port(adap, i)
		enable_txq_db(&adap->sge.ctrlq[i].q);
}

static void sync_txq_pidx(struct adapter *adap, struct sge_txq *q)
{
	u16 hw_pidx, hw_cidx;
	int ret;

	spin_lock_bh(&q->db_lock);
	ret = read_eq_indices(adap, (u16)q->cntxt_id, &hw_pidx, &hw_cidx);
	if (ret)
		goto out;
	if (q->db_pidx != hw_pidx) {
		u16 delta;

		if (q->db_pidx >= hw_pidx)
			delta = q->db_pidx - hw_pidx;
		else
			delta = q->size - hw_pidx + q->db_pidx;
		wmb();
		t4_write_reg(adap, MYPF_REG(SGE_PF_KDOORBELL),
			     QID(q->cntxt_id) | PIDX(delta));
	}
out:
	q->db_disabled = 0;
	spin_unlock_bh(&q->db_lock);
	if (ret)
		CH_WARN(adap, "DB drop recovery failed.\n");
}
static void recover_all_queues(struct adapter *adap)
{
	int i;

	for_each_ethrxq(&adap->sge, i)
		sync_txq_pidx(adap, &adap->sge.ethtxq[i].q);
	for_each_ofldrxq(&adap->sge, i)
		sync_txq_pidx(adap, &adap->sge.ofldtxq[i].q);
	for_each_port(adap, i)
		sync_txq_pidx(adap, &adap->sge.ctrlq[i].q);
}

static void notify_rdma_uld(struct adapter *adap, enum cxgb4_control cmd)
{
	mutex_lock(&uld_mutex);
	if (adap->uld_handle[CXGB4_ULD_RDMA])
		ulds[CXGB4_ULD_RDMA].control(adap->uld_handle[CXGB4_ULD_RDMA],
				cmd);
	mutex_unlock(&uld_mutex);
}

static void process_db_full(struct work_struct *work)
{
	struct adapter *adap;

	adap = container_of(work, struct adapter, db_full_task);

	notify_rdma_uld(adap, CXGB4_CONTROL_DB_FULL);
	drain_db_fifo(adap, dbfifo_drain_delay);
	t4_set_reg_field(adap, SGE_INT_ENABLE3,
			 DBFIFO_HP_INT | DBFIFO_LP_INT,
			 DBFIFO_HP_INT | DBFIFO_LP_INT);
	notify_rdma_uld(adap, CXGB4_CONTROL_DB_EMPTY);
}

static void process_db_drop(struct work_struct *work)
{
	struct adapter *adap;

	adap = container_of(work, struct adapter, db_drop_task);

	t4_set_reg_field(adap, A_SGE_DOORBELL_CONTROL, F_DROPPED_DB, 0);
	disable_dbs(adap);
	notify_rdma_uld(adap, CXGB4_CONTROL_DB_DROP);
	drain_db_fifo(adap, 1);
	recover_all_queues(adap);
	enable_dbs(adap);
}

void t4_db_full(struct adapter *adap)
{
	t4_set_reg_field(adap, SGE_INT_ENABLE3,
			 DBFIFO_HP_INT | DBFIFO_LP_INT, 0);
	queue_work(workq, &adap->db_full_task);
}

void t4_db_dropped(struct adapter *adap)
{
	queue_work(workq, &adap->db_drop_task);
}

static void uld_attach(struct adapter *adap, unsigned int uld)
{
	void *handle;
	struct cxgb4_lld_info lli;

	lli.pdev = adap->pdev;
	lli.l2t = adap->l2t;
	lli.tids = &adap->tids;
	lli.ports = adap->port;
	lli.vr = &adap->vres;
	lli.mtus = adap->params.mtus;
	if (uld == CXGB4_ULD_RDMA) {
		lli.rxq_ids = adap->sge.rdma_rxq;
		lli.nrxq = adap->sge.rdmaqs;
	} else if (uld == CXGB4_ULD_ISCSI) {
		lli.rxq_ids = adap->sge.ofld_rxq;
		lli.nrxq = adap->sge.ofldqsets;
	}
	lli.ntxq = adap->sge.ofldqsets;
	lli.nchan = adap->params.nports;
	lli.nports = adap->params.nports;
	lli.wr_cred = adap->params.ofldq_wr_cred;
	lli.adapter_type = adap->params.rev;
	lli.iscsi_iolen = MAXRXDATA_GET(t4_read_reg(adap, TP_PARA_REG2));
	lli.udb_density = 1 << QUEUESPERPAGEPF0_GET(
			t4_read_reg(adap, SGE_EGRESS_QUEUES_PER_PAGE_PF) >>
			(adap->fn * 4));
	lli.ucq_density = 1 << QUEUESPERPAGEPF0_GET(
			t4_read_reg(adap, SGE_INGRESS_QUEUES_PER_PAGE_PF) >>
			(adap->fn * 4));
	lli.gts_reg = adap->regs + MYPF_REG(SGE_PF_GTS);
	lli.db_reg = adap->regs + MYPF_REG(SGE_PF_KDOORBELL);
	lli.fw_vers = adap->params.fw_vers;
	lli.dbfifo_int_thresh = dbfifo_int_thresh;

	handle = ulds[uld].add(&lli);
	if (IS_ERR(handle)) {
		dev_warn(adap->pdev_dev,
			 "could not attach to the %s driver, error %ld\n",
			 uld_str[uld], PTR_ERR(handle));
		return;
	}

	adap->uld_handle[uld] = handle;

	if (!netevent_registered) {
		register_netevent_notifier(&cxgb4_netevent_nb);
		netevent_registered = true;
	}

	if (adap->flags & FULL_INIT_DONE)
		ulds[uld].state_change(handle, CXGB4_STATE_UP);
}

static void attach_ulds(struct adapter *adap)
{
	unsigned int i;

	mutex_lock(&uld_mutex);
	list_add_tail(&adap->list_node, &adapter_list);
	for (i = 0; i < CXGB4_ULD_MAX; i++)
		if (ulds[i].add)
			uld_attach(adap, i);
	mutex_unlock(&uld_mutex);
}

static void detach_ulds(struct adapter *adap)
{
	unsigned int i;

	mutex_lock(&uld_mutex);
	list_del(&adap->list_node);
	for (i = 0; i < CXGB4_ULD_MAX; i++)
		if (adap->uld_handle[i]) {
			ulds[i].state_change(adap->uld_handle[i],
					     CXGB4_STATE_DETACH);
			adap->uld_handle[i] = NULL;
		}
	if (netevent_registered && list_empty(&adapter_list)) {
		unregister_netevent_notifier(&cxgb4_netevent_nb);
		netevent_registered = false;
	}
	mutex_unlock(&uld_mutex);
}

static void notify_ulds(struct adapter *adap, enum cxgb4_state new_state)
{
	unsigned int i;

	mutex_lock(&uld_mutex);
	for (i = 0; i < CXGB4_ULD_MAX; i++)
		if (adap->uld_handle[i])
			ulds[i].state_change(adap->uld_handle[i], new_state);
	mutex_unlock(&uld_mutex);
}

/**
 *	cxgb4_register_uld - register an upper-layer driver
 *	@type: the ULD type
 *	@p: the ULD methods
 *
 *	Registers an upper-layer driver with this driver and notifies the ULD
 *	about any presently available devices that support its type.  Returns
 *	%-EBUSY if a ULD of the same type is already registered.
 */
int cxgb4_register_uld(enum cxgb4_uld type, const struct cxgb4_uld_info *p)
{
	int ret = 0;
	struct adapter *adap;

	if (type >= CXGB4_ULD_MAX)
		return -EINVAL;
	mutex_lock(&uld_mutex);
	if (ulds[type].add) {
		ret = -EBUSY;
		goto out;
	}
	ulds[type] = *p;
	list_for_each_entry(adap, &adapter_list, list_node)
		uld_attach(adap, type);
out:	mutex_unlock(&uld_mutex);
	return ret;
}
EXPORT_SYMBOL(cxgb4_register_uld);

/**
 *	cxgb4_unregister_uld - unregister an upper-layer driver
 *	@type: the ULD type
 *
 *	Unregisters an existing upper-layer driver.
 */
int cxgb4_unregister_uld(enum cxgb4_uld type)
{
	struct adapter *adap;

	if (type >= CXGB4_ULD_MAX)
		return -EINVAL;
	mutex_lock(&uld_mutex);
	list_for_each_entry(adap, &adapter_list, list_node)
		adap->uld_handle[type] = NULL;
	ulds[type].add = NULL;
	mutex_unlock(&uld_mutex);
	return 0;
}
EXPORT_SYMBOL(cxgb4_unregister_uld);

/**
 *	cxgb_up - enable the adapter
 *	@adap: adapter being enabled
 *
 *	Called when the first port is enabled, this function performs the
 *	actions necessary to make an adapter operational, such as completing
 *	the initialization of HW modules, and enabling interrupts.
 *
 *	Must be called with the rtnl lock held.
 */
static int cxgb_up(struct adapter *adap)
{
	int err;

	err = setup_sge_queues(adap);
	if (err)
		goto out;
	err = setup_rss(adap);
	if (err)
		goto freeq;

	if (adap->flags & USING_MSIX) {
		name_msix_vecs(adap);
		err = request_irq(adap->msix_info[0].vec, t4_nondata_intr, 0,
				  adap->msix_info[0].desc, adap);
		if (err)
			goto irq_err;

		err = request_msix_queue_irqs(adap);
		if (err) {
			free_irq(adap->msix_info[0].vec, adap);
			goto irq_err;
		}
	} else {
		err = request_irq(adap->pdev->irq, t4_intr_handler(adap),
				  (adap->flags & USING_MSI) ? 0 : IRQF_SHARED,
				  adap->port[0]->name, adap);
		if (err)
			goto irq_err;
	}
	enable_rx(adap);
	t4_sge_start(adap);
	t4_intr_enable(adap);
	adap->flags |= FULL_INIT_DONE;
	notify_ulds(adap, CXGB4_STATE_UP);
 out:
	return err;
 irq_err:
	dev_err(adap->pdev_dev, "request_irq failed, err %d\n", err);
 freeq:
	t4_free_sge_resources(adap);
	goto out;
}

static void cxgb_down(struct adapter *adapter)
{
	t4_intr_disable(adapter);
	cancel_work_sync(&adapter->tid_release_task);
	cancel_work_sync(&adapter->db_full_task);
	cancel_work_sync(&adapter->db_drop_task);
	adapter->tid_release_task_busy = false;
	adapter->tid_release_head = NULL;

	if (adapter->flags & USING_MSIX) {
		free_msix_queue_irqs(adapter);
		free_irq(adapter->msix_info[0].vec, adapter);
	} else
		free_irq(adapter->pdev->irq, adapter);
	quiesce_rx(adapter);
	t4_sge_stop(adapter);
	t4_free_sge_resources(adapter);
	adapter->flags &= ~FULL_INIT_DONE;
}

/*
 * net_device operations
 */
static int cxgb_open(struct net_device *dev)
{
	int err;
	struct port_info *pi = netdev_priv(dev);
	struct adapter *adapter = pi->adapter;

	netif_carrier_off(dev);

	if (!(adapter->flags & FULL_INIT_DONE)) {
		err = cxgb_up(adapter);
		if (err < 0)
			return err;
	}

	err = link_start(dev);
	if (!err)
		netif_tx_start_all_queues(dev);
	return err;
}

static int cxgb_close(struct net_device *dev)
{
	struct port_info *pi = netdev_priv(dev);
	struct adapter *adapter = pi->adapter;

	netif_tx_stop_all_queues(dev);
	netif_carrier_off(dev);
	return t4_enable_vi(adapter, adapter->fn, pi->viid, false, false);
}

/* Return an error number if the indicated filter isn't writable ...
 */
static int writable_filter(struct filter_entry *f)
{
	if (f->locked)
		return -EPERM;
	if (f->pending)
		return -EBUSY;

	return 0;
}

/* Delete the filter at the specified index (if valid).  The checks for all
 * the common problems with doing this like the filter being locked, currently
 * pending in another operation, etc.
 */
static int delete_filter(struct adapter *adapter, unsigned int fidx)
{
	struct filter_entry *f;
	int ret;

	if (fidx >= adapter->tids.nftids)
		return -EINVAL;

	f = &adapter->tids.ftid_tab[fidx];
	ret = writable_filter(f);
	if (ret)
		return ret;
	if (f->valid)
		return del_filter_wr(adapter, fidx);

	return 0;
}

static struct rtnl_link_stats64 *cxgb_get_stats(struct net_device *dev,
						struct rtnl_link_stats64 *ns)
{
	struct port_stats stats;
	struct port_info *p = netdev_priv(dev);
	struct adapter *adapter = p->adapter;

	spin_lock(&adapter->stats_lock);
	t4_get_port_stats(adapter, p->tx_chan, &stats);
	spin_unlock(&adapter->stats_lock);

	ns->tx_bytes   = stats.tx_octets;
	ns->tx_packets = stats.tx_frames;
	ns->rx_bytes   = stats.rx_octets;
	ns->rx_packets = stats.rx_frames;
	ns->multicast  = stats.rx_mcast_frames;

	/* detailed rx_errors */
	ns->rx_length_errors = stats.rx_jabber + stats.rx_too_long +
			       stats.rx_runt;
	ns->rx_over_errors   = 0;
	ns->rx_crc_errors    = stats.rx_fcs_err;
	ns->rx_frame_errors  = stats.rx_symbol_err;
	ns->rx_fifo_errors   = stats.rx_ovflow0 + stats.rx_ovflow1 +
			       stats.rx_ovflow2 + stats.rx_ovflow3 +
			       stats.rx_trunc0 + stats.rx_trunc1 +
			       stats.rx_trunc2 + stats.rx_trunc3;
	ns->rx_missed_errors = 0;

	/* detailed tx_errors */
	ns->tx_aborted_errors   = 0;
	ns->tx_carrier_errors   = 0;
	ns->tx_fifo_errors      = 0;
	ns->tx_heartbeat_errors = 0;
	ns->tx_window_errors    = 0;

	ns->tx_errors = stats.tx_error_frames;
	ns->rx_errors = stats.rx_symbol_err + stats.rx_fcs_err +
		ns->rx_length_errors + stats.rx_len_err + ns->rx_fifo_errors;
	return ns;
}

static int cxgb_ioctl(struct net_device *dev, struct ifreq *req, int cmd)
{
	unsigned int mbox;
	int ret = 0, prtad, devad;
	struct port_info *pi = netdev_priv(dev);
	struct mii_ioctl_data *data = (struct mii_ioctl_data *)&req->ifr_data;

	switch (cmd) {
	case SIOCGMIIPHY:
		if (pi->mdio_addr < 0)
			return -EOPNOTSUPP;
		data->phy_id = pi->mdio_addr;
		break;
	case SIOCGMIIREG:
	case SIOCSMIIREG:
		if (mdio_phy_id_is_c45(data->phy_id)) {
			prtad = mdio_phy_id_prtad(data->phy_id);
			devad = mdio_phy_id_devad(data->phy_id);
		} else if (data->phy_id < 32) {
			prtad = data->phy_id;
			devad = 0;
			data->reg_num &= 0x1f;
		} else
			return -EINVAL;

		mbox = pi->adapter->fn;
		if (cmd == SIOCGMIIREG)
			ret = t4_mdio_rd(pi->adapter, mbox, prtad, devad,
					 data->reg_num, &data->val_out);
		else
			ret = t4_mdio_wr(pi->adapter, mbox, prtad, devad,
					 data->reg_num, data->val_in);
		break;
	default:
		return -EOPNOTSUPP;
	}
	return ret;
}

static void cxgb_set_rxmode(struct net_device *dev)
{
	/* unfortunately we can't return errors to the stack */
	set_rxmode(dev, -1, false);
}

static int cxgb_change_mtu(struct net_device *dev, int new_mtu)
{
	int ret;
	struct port_info *pi = netdev_priv(dev);

	if (new_mtu < 81 || new_mtu > MAX_MTU)         /* accommodate SACK */
		return -EINVAL;
	ret = t4_set_rxmode(pi->adapter, pi->adapter->fn, pi->viid, new_mtu, -1,
			    -1, -1, -1, true);
	if (!ret)
		dev->mtu = new_mtu;
	return ret;
}

static int cxgb_set_mac_addr(struct net_device *dev, void *p)
{
	int ret;
	struct sockaddr *addr = p;
	struct port_info *pi = netdev_priv(dev);

	if (!is_valid_ether_addr(addr->sa_data))
		return -EADDRNOTAVAIL;

	ret = t4_change_mac(pi->adapter, pi->adapter->fn, pi->viid,
			    pi->xact_addr_filt, addr->sa_data, true, true);
	if (ret < 0)
		return ret;

	memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
	pi->xact_addr_filt = ret;
	return 0;
}

#ifdef CONFIG_NET_POLL_CONTROLLER
static void cxgb_netpoll(struct net_device *dev)
{
	struct port_info *pi = netdev_priv(dev);
	struct adapter *adap = pi->adapter;

	if (adap->flags & USING_MSIX) {
		int i;
		struct sge_eth_rxq *rx = &adap->sge.ethrxq[pi->first_qset];

		for (i = pi->nqsets; i; i--, rx++)
			t4_sge_intr_msix(0, &rx->rspq);
	} else
		t4_intr_handler(adap)(0, adap);
}
#endif

static const struct net_device_ops cxgb4_netdev_ops = {
	.ndo_open             = cxgb_open,
	.ndo_stop             = cxgb_close,
	.ndo_start_xmit       = t4_eth_xmit,
	.ndo_get_stats64      = cxgb_get_stats,
	.ndo_set_rx_mode      = cxgb_set_rxmode,
	.ndo_set_mac_address  = cxgb_set_mac_addr,
	.ndo_set_features     = cxgb_set_features,
	.ndo_validate_addr    = eth_validate_addr,
	.ndo_do_ioctl         = cxgb_ioctl,
	.ndo_change_mtu       = cxgb_change_mtu,
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller  = cxgb_netpoll,
#endif
};

void t4_fatal_err(struct adapter *adap)
{
	t4_set_reg_field(adap, SGE_CONTROL, GLOBALENABLE, 0);
	t4_intr_disable(adap);
	dev_alert(adap->pdev_dev, "encountered fatal error, adapter stopped\n");
}

static void setup_memwin(struct adapter *adap)
{
	u32 bar0;

	bar0 = pci_resource_start(adap->pdev, 0);  /* truncation intentional */
	t4_write_reg(adap, PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 0),
		     (bar0 + MEMWIN0_BASE) | BIR(0) |
		     WINDOW(ilog2(MEMWIN0_APERTURE) - 10));
	t4_write_reg(adap, PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 1),
		     (bar0 + MEMWIN1_BASE) | BIR(0) |
		     WINDOW(ilog2(MEMWIN1_APERTURE) - 10));
	t4_write_reg(adap, PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 2),
		     (bar0 + MEMWIN2_BASE) | BIR(0) |
		     WINDOW(ilog2(MEMWIN2_APERTURE) - 10));
}

static void setup_memwin_rdma(struct adapter *adap)
{
	if (adap->vres.ocq.size) {
		unsigned int start, sz_kb;

		start = pci_resource_start(adap->pdev, 2) +
			OCQ_WIN_OFFSET(adap->pdev, &adap->vres);
		sz_kb = roundup_pow_of_two(adap->vres.ocq.size) >> 10;
		t4_write_reg(adap,
			     PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_BASE_WIN, 3),
			     start | BIR(1) | WINDOW(ilog2(sz_kb)));
		t4_write_reg(adap,
			     PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET, 3),
			     adap->vres.ocq.start);
		t4_read_reg(adap,
			    PCIE_MEM_ACCESS_REG(PCIE_MEM_ACCESS_OFFSET, 3));
	}
}

static int adap_init1(struct adapter *adap, struct fw_caps_config_cmd *c)
{
	u32 v;
	int ret;

	/* get device capabilities */
	memset(c, 0, sizeof(*c));
	c->op_to_write = htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
			       FW_CMD_REQUEST | FW_CMD_READ);
	c->cfvalid_to_len16 = htonl(FW_LEN16(*c));
	ret = t4_wr_mbox(adap, adap->fn, c, sizeof(*c), c);
	if (ret < 0)
		return ret;

	/* select capabilities we'll be using */
	if (c->niccaps & htons(FW_CAPS_CONFIG_NIC_VM)) {
		if (!vf_acls)
			c->niccaps ^= htons(FW_CAPS_CONFIG_NIC_VM);
		else
			c->niccaps = htons(FW_CAPS_CONFIG_NIC_VM);
	} else if (vf_acls) {
		dev_err(adap->pdev_dev, "virtualization ACLs not supported");
		return ret;
	}
	c->op_to_write = htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
			       FW_CMD_REQUEST | FW_CMD_WRITE);
	ret = t4_wr_mbox(adap, adap->fn, c, sizeof(*c), NULL);
	if (ret < 0)
		return ret;

	ret = t4_config_glbl_rss(adap, adap->fn,
				 FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL,
				 FW_RSS_GLB_CONFIG_CMD_TNLMAPEN |
				 FW_RSS_GLB_CONFIG_CMD_TNLALLLKP);
	if (ret < 0)
		return ret;

	ret = t4_cfg_pfvf(adap, adap->fn, adap->fn, 0, MAX_EGRQ, 64, MAX_INGQ,
			  0, 0, 4, 0xf, 0xf, 16, FW_CMD_CAP_PF, FW_CMD_CAP_PF);
	if (ret < 0)
		return ret;

	t4_sge_init(adap);

	/* tweak some settings */
	t4_write_reg(adap, TP_SHIFT_CNT, 0x64f8849);
	t4_write_reg(adap, ULP_RX_TDDP_PSZ, HPZ0(PAGE_SHIFT - 12));
	t4_write_reg(adap, TP_PIO_ADDR, TP_INGRESS_CONFIG);
	v = t4_read_reg(adap, TP_PIO_DATA);
	t4_write_reg(adap, TP_PIO_DATA, v & ~CSUM_HAS_PSEUDO_HDR);

	/* get basic stuff going */
	return t4_early_init(adap, adap->fn);
}

/*
 * Max # of ATIDs.  The absolute HW max is 16K but we keep it lower.
 */
#define MAX_ATIDS 8192U

/*
 * Phase 0 of initialization: contact FW, obtain config, perform basic init.
 *
 * If the firmware we're dealing with has Configuration File support, then
 * we use that to perform all configuration
 */

/*
 * Tweak configuration based on module parameters, etc.  Most of these have
 * defaults assigned to them by Firmware Configuration Files (if we're using
 * them) but need to be explicitly set if we're using hard-coded
 * initialization.  But even in the case of using Firmware Configuration
 * Files, we'd like to expose the ability to change these via module
 * parameters so these are essentially common tweaks/settings for
 * Configuration Files and hard-coded initialization ...
 */
static int adap_init0_tweaks(struct adapter *adapter)
{
	/*
	 * Fix up various Host-Dependent Parameters like Page Size, Cache
	 * Line Size, etc.  The firmware default is for a 4KB Page Size and
	 * 64B Cache Line Size ...
	 */
	t4_fixup_host_params(adapter, PAGE_SIZE, L1_CACHE_BYTES);

	/*
	 * Process module parameters which affect early initialization.
	 */
	if (rx_dma_offset != 2 && rx_dma_offset != 0) {
		dev_err(&adapter->pdev->dev,
			"Ignoring illegal rx_dma_offset=%d, using 2\n",
			rx_dma_offset);
		rx_dma_offset = 2;
	}
	t4_set_reg_field(adapter, SGE_CONTROL,
			 PKTSHIFT_MASK,
			 PKTSHIFT(rx_dma_offset));

	/*
	 * Don't include the "IP Pseudo Header" in CPL_RX_PKT checksums: Linux
	 * adds the pseudo header itself.
	 */
	t4_tp_wr_bits_indirect(adapter, TP_INGRESS_CONFIG,
			       CSUM_HAS_PSEUDO_HDR, 0);

	return 0;
}

/*
 * Attempt to initialize the adapter via a Firmware Configuration File.
 */
static int adap_init0_config(struct adapter *adapter, int reset)
{
	struct fw_caps_config_cmd caps_cmd;
	const struct firmware *cf;
	unsigned long mtype = 0, maddr = 0;
	u32 finiver, finicsum, cfcsum;
	int ret, using_flash;

	/*
	 * Reset device if necessary.
	 */
	if (reset) {
		ret = t4_fw_reset(adapter, adapter->mbox,
				  PIORSTMODE | PIORST);
		if (ret < 0)
			goto bye;
	}

	/*
	 * If we have a T4 configuration file under /lib/firmware/cxgb4/,
	 * then use that.  Otherwise, use the configuration file stored
	 * in the adapter flash ...
	 */
	ret = request_firmware(&cf, FW_CFNAME, adapter->pdev_dev);
	if (ret < 0) {
		using_flash = 1;
		mtype = FW_MEMTYPE_CF_FLASH;
		maddr = t4_flash_cfg_addr(adapter);
	} else {
		u32 params[7], val[7];

		using_flash = 0;
		if (cf->size >= FLASH_CFG_MAX_SIZE)
			ret = -ENOMEM;
		else {
			params[0] = (FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) |
			     FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_CF));
			ret = t4_query_params(adapter, adapter->mbox,
					      adapter->fn, 0, 1, params, val);
			if (ret == 0) {
				/*
				 * For t4_memory_write() below addresses and
				 * sizes have to be in terms of multiples of 4
				 * bytes.  So, if the Configuration File isn't
				 * a multiple of 4 bytes in length we'll have
				 * to write that out separately since we can't
				 * guarantee that the bytes following the
				 * residual byte in the buffer returned by
				 * request_firmware() are zeroed out ...
				 */
				size_t resid = cf->size & 0x3;
				size_t size = cf->size & ~0x3;
				__be32 *data = (__be32 *)cf->data;

				mtype = FW_PARAMS_PARAM_Y_GET(val[0]);
				maddr = FW_PARAMS_PARAM_Z_GET(val[0]) << 16;

				ret = t4_memory_write(adapter, mtype, maddr,
						      size, data);
				if (ret == 0 && resid != 0) {
					union {
						__be32 word;
						char buf[4];
					} last;
					int i;

					last.word = data[size >> 2];
					for (i = resid; i < 4; i++)
						last.buf[i] = 0;
					ret = t4_memory_write(adapter, mtype,
							      maddr + size,
							      4, &last.word);
				}
			}
		}

		release_firmware(cf);
		if (ret)
			goto bye;
	}

	/*
	 * Issue a Capability Configuration command to the firmware to get it
	 * to parse the Configuration File.  We don't use t4_fw_config_file()
	 * because we want the ability to modify various features after we've
	 * processed the configuration file ...
	 */
	memset(&caps_cmd, 0, sizeof(caps_cmd));
	caps_cmd.op_to_write =
		htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
		      FW_CMD_REQUEST |
		      FW_CMD_READ);
	caps_cmd.cfvalid_to_len16 =
		htonl(FW_CAPS_CONFIG_CMD_CFVALID |
		      FW_CAPS_CONFIG_CMD_MEMTYPE_CF(mtype) |
		      FW_CAPS_CONFIG_CMD_MEMADDR64K_CF(maddr >> 16) |
		      FW_LEN16(caps_cmd));
	ret = t4_wr_mbox(adapter, adapter->mbox, &caps_cmd, sizeof(caps_cmd),
			 &caps_cmd);
	if (ret < 0)
		goto bye;

	finiver = ntohl(caps_cmd.finiver);
	finicsum = ntohl(caps_cmd.finicsum);
	cfcsum = ntohl(caps_cmd.cfcsum);
	if (finicsum != cfcsum)
		dev_warn(adapter->pdev_dev, "Configuration File checksum "\
			 "mismatch: [fini] csum=%#x, computed csum=%#x\n",
			 finicsum, cfcsum);

	/*
	 * And now tell the firmware to use the configuration we just loaded.
	 */
	caps_cmd.op_to_write =
		htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
		      FW_CMD_REQUEST |
		      FW_CMD_WRITE);
	caps_cmd.cfvalid_to_len16 = htonl(FW_LEN16(caps_cmd));
	ret = t4_wr_mbox(adapter, adapter->mbox, &caps_cmd, sizeof(caps_cmd),
			 NULL);
	if (ret < 0)
		goto bye;

	/*
	 * Tweak configuration based on system architecture, module
	 * parameters, etc.
	 */
	ret = adap_init0_tweaks(adapter);
	if (ret < 0)
		goto bye;

	/*
	 * And finally tell the firmware to initialize itself using the
	 * parameters from the Configuration File.
	 */
	ret = t4_fw_initialize(adapter, adapter->mbox);
	if (ret < 0)
		goto bye;

	/*
	 * Return successfully and note that we're operating with parameters
	 * not supplied by the driver, rather than from hard-wired
	 * initialization constants burried in the driver.
	 */
	adapter->flags |= USING_SOFT_PARAMS;
	dev_info(adapter->pdev_dev, "Successfully configured using Firmware "\
		 "Configuration File %s, version %#x, computed checksum %#x\n",
		 (using_flash
		  ? "in device FLASH"
		  : "/lib/firmware/" FW_CFNAME),
		 finiver, cfcsum);
	return 0;

	/*
	 * Something bad happened.  Return the error ...  (If the "error"
	 * is that there's no Configuration File on the adapter we don't
	 * want to issue a warning since this is fairly common.)
	 */
bye:
	if (ret != -ENOENT)
		dev_warn(adapter->pdev_dev, "Configuration file error %d\n",
			 -ret);
	return ret;
}

/*
 * Attempt to initialize the adapter via hard-coded, driver supplied
 * parameters ...
 */
static int adap_init0_no_config(struct adapter *adapter, int reset)
{
	struct sge *s = &adapter->sge;
	struct fw_caps_config_cmd caps_cmd;
	u32 v;
	int i, ret;

	/*
	 * Reset device if necessary
	 */
	if (reset) {
		ret = t4_fw_reset(adapter, adapter->mbox,
				  PIORSTMODE | PIORST);
		if (ret < 0)
			goto bye;
	}

	/*
	 * Get device capabilities and select which we'll be using.
	 */
	memset(&caps_cmd, 0, sizeof(caps_cmd));
	caps_cmd.op_to_write = htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
				     FW_CMD_REQUEST | FW_CMD_READ);
	caps_cmd.cfvalid_to_len16 = htonl(FW_LEN16(caps_cmd));
	ret = t4_wr_mbox(adapter, adapter->mbox, &caps_cmd, sizeof(caps_cmd),
			 &caps_cmd);
	if (ret < 0)
		goto bye;

	if (caps_cmd.niccaps & htons(FW_CAPS_CONFIG_NIC_VM)) {
		if (!vf_acls)
			caps_cmd.niccaps ^= htons(FW_CAPS_CONFIG_NIC_VM);
		else
			caps_cmd.niccaps = htons(FW_CAPS_CONFIG_NIC_VM);
	} else if (vf_acls) {
		dev_err(adapter->pdev_dev, "virtualization ACLs not supported");
		goto bye;
	}
	caps_cmd.op_to_write = htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
			      FW_CMD_REQUEST | FW_CMD_WRITE);
	ret = t4_wr_mbox(adapter, adapter->mbox, &caps_cmd, sizeof(caps_cmd),
			 NULL);
	if (ret < 0)
		goto bye;

	/*
	 * Tweak configuration based on system architecture, module
	 * parameters, etc.
	 */
	ret = adap_init0_tweaks(adapter);
	if (ret < 0)
		goto bye;

	/*
	 * Select RSS Global Mode we want to use.  We use "Basic Virtual"
	 * mode which maps each Virtual Interface to its own section of
	 * the RSS Table and we turn on all map and hash enables ...
	 */
	adapter->flags |= RSS_TNLALLLOOKUP;
	ret = t4_config_glbl_rss(adapter, adapter->mbox,
				 FW_RSS_GLB_CONFIG_CMD_MODE_BASICVIRTUAL,
				 FW_RSS_GLB_CONFIG_CMD_TNLMAPEN |
				 FW_RSS_GLB_CONFIG_CMD_HASHTOEPLITZ |
				 ((adapter->flags & RSS_TNLALLLOOKUP) ?
					FW_RSS_GLB_CONFIG_CMD_TNLALLLKP : 0));
	if (ret < 0)
		goto bye;

	/*
	 * Set up our own fundamental resource provisioning ...
	 */
	ret = t4_cfg_pfvf(adapter, adapter->mbox, adapter->fn, 0,
			  PFRES_NEQ, PFRES_NETHCTRL,
			  PFRES_NIQFLINT, PFRES_NIQ,
			  PFRES_TC, PFRES_NVI,
			  FW_PFVF_CMD_CMASK_MASK,
			  pfvfres_pmask(adapter, adapter->fn, 0),
			  PFRES_NEXACTF,
			  PFRES_R_CAPS, PFRES_WX_CAPS);
	if (ret < 0)
		goto bye;

	/*
	 * Perform low level SGE initialization.  We need to do this before we
	 * send the firmware the INITIALIZE command because that will cause
	 * any other PF Drivers which are waiting for the Master
	 * Initialization to proceed forward.
	 */
	for (i = 0; i < SGE_NTIMERS - 1; i++)
		s->timer_val[i] = min(intr_holdoff[i], MAX_SGE_TIMERVAL);
	s->timer_val[SGE_NTIMERS - 1] = MAX_SGE_TIMERVAL;
	s->counter_val[0] = 1;
	for (i = 1; i < SGE_NCOUNTERS; i++)
		s->counter_val[i] = min(intr_cnt[i - 1],
					THRESHOLD_0_GET(THRESHOLD_0_MASK));
	t4_sge_init(adapter);

#ifdef CONFIG_PCI_IOV
	/*
	 * Provision resource limits for Virtual Functions.  We currently
	 * grant them all the same static resource limits except for the Port
	 * Access Rights Mask which we're assigning based on the PF.  All of
	 * the static provisioning stuff for both the PF and VF really needs
	 * to be managed in a persistent manner for each device which the
	 * firmware controls.
	 */
	{
		int pf, vf;

		for (pf = 0; pf < ARRAY_SIZE(num_vf); pf++) {
			if (num_vf[pf] <= 0)
				continue;

			/* VF numbering starts at 1! */
			for (vf = 1; vf <= num_vf[pf]; vf++) {
				ret = t4_cfg_pfvf(adapter, adapter->mbox,
						  pf, vf,
						  VFRES_NEQ, VFRES_NETHCTRL,
						  VFRES_NIQFLINT, VFRES_NIQ,
						  VFRES_TC, VFRES_NVI,
						  FW_PFVF_CMD_CMASK_GET(
						  FW_PFVF_CMD_CMASK_MASK),
						  pfvfres_pmask(
						  adapter, pf, vf),
						  VFRES_NEXACTF,
						  VFRES_R_CAPS, VFRES_WX_CAPS);
				if (ret < 0)
					dev_warn(adapter->pdev_dev,
						 "failed to "\
						 "provision pf/vf=%d/%d; "
						 "err=%d\n", pf, vf, ret);
			}
		}
	}
#endif

	/*
	 * Set up the default filter mode.  Later we'll want to implement this
	 * via a firmware command, etc. ...  This needs to be done before the
	 * firmare initialization command ...  If the selected set of fields
	 * isn't equal to the default value, we'll need to make sure that the
	 * field selections will fit in the 36-bit budget.
	 */
	if (tp_vlan_pri_map != TP_VLAN_PRI_MAP_DEFAULT) {
		int j, bits = 0;

		for (j = TP_VLAN_PRI_MAP_FIRST; j <= TP_VLAN_PRI_MAP_LAST; j++)
			switch (tp_vlan_pri_map & (1 << j)) {
			case 0:
				/* compressed filter field not enabled */
				break;
			case FCOE_MASK:
				bits +=  1;
				break;
			case PORT_MASK:
				bits +=  3;
				break;
			case VNIC_ID_MASK:
				bits += 17;
				break;
			case VLAN_MASK:
				bits += 17;
				break;
			case TOS_MASK:
				bits +=  8;
				break;
			case PROTOCOL_MASK:
				bits +=  8;
				break;
			case ETHERTYPE_MASK:
				bits += 16;
				break;
			case MACMATCH_MASK:
				bits +=  9;
				break;
			case MPSHITTYPE_MASK:
				bits +=  3;
				break;
			case FRAGMENTATION_MASK:
				bits +=  1;
				break;
			}

		if (bits > 36) {
			dev_err(adapter->pdev_dev,
				"tp_vlan_pri_map=%#x needs %d bits > 36;"\
				" using %#x\n", tp_vlan_pri_map, bits,
				TP_VLAN_PRI_MAP_DEFAULT);
			tp_vlan_pri_map = TP_VLAN_PRI_MAP_DEFAULT;
		}
	}
	v = tp_vlan_pri_map;
	t4_write_indirect(adapter, TP_PIO_ADDR, TP_PIO_DATA,
			  &v, 1, TP_VLAN_PRI_MAP);

	/*
	 * We need Five Tuple Lookup mode to be set in TP_GLOBAL_CONFIG order
	 * to support any of the compressed filter fields above.  Newer
	 * versions of the firmware do this automatically but it doesn't hurt
	 * to set it here.  Meanwhile, we do _not_ need to set Lookup Every
	 * Packet in TP_INGRESS_CONFIG to support matching non-TCP packets
	 * since the firmware automatically turns this on and off when we have
	 * a non-zero number of filters active (since it does have a
	 * performance impact).
	 */
	if (tp_vlan_pri_map)
		t4_set_reg_field(adapter, TP_GLOBAL_CONFIG,
				 FIVETUPLELOOKUP_MASK,
				 FIVETUPLELOOKUP_MASK);

	/*
	 * Tweak some settings.
	 */
	t4_write_reg(adapter, TP_SHIFT_CNT, SYNSHIFTMAX(6) |
		     RXTSHIFTMAXR1(4) | RXTSHIFTMAXR2(15) |
		     PERSHIFTBACKOFFMAX(8) | PERSHIFTMAX(8) |
		     KEEPALIVEMAXR1(4) | KEEPALIVEMAXR2(9));

	/*
	 * Get basic stuff going by issuing the Firmware Initialize command.
	 * Note that this _must_ be after all PFVF commands ...
	 */
	ret = t4_fw_initialize(adapter, adapter->mbox);
	if (ret < 0)
		goto bye;

	/*
	 * Return successfully!
	 */
	dev_info(adapter->pdev_dev, "Successfully configured using built-in "\
		 "driver parameters\n");
	return 0;

	/*
	 * Something bad happened.  Return the error ...
	 */
bye:
	return ret;
}

/*
 * Phase 0 of initialization: contact FW, obtain config, perform basic init.
 */
static int adap_init0(struct adapter *adap)
{
	int ret;
	u32 v, port_vec;
	enum dev_state state;
	u32 params[7], val[7];
	struct fw_caps_config_cmd caps_cmd;
	int reset = 1, j;

	/*
	 * Contact FW, advertising Master capability (and potentially forcing
	 * ourselves as the Master PF if our module parameter force_init is
	 * set).
	 */
	ret = t4_fw_hello(adap, adap->mbox, adap->fn,
			  force_init ? MASTER_MUST : MASTER_MAY,
			  &state);
	if (ret < 0) {
		dev_err(adap->pdev_dev, "could not connect to FW, error %d\n",
			ret);
		return ret;
	}
	if (ret == adap->mbox)
		adap->flags |= MASTER_PF;
	if (force_init && state == DEV_STATE_INIT)
		state = DEV_STATE_UNINIT;

	/*
	 * If we're the Master PF Driver and the device is uninitialized,
	 * then let's consider upgrading the firmware ...  (We always want
	 * to check the firmware version number in order to A. get it for
	 * later reporting and B. to warn if the currently loaded firmware
	 * is excessively mismatched relative to the driver.)
	 */
	ret = t4_check_fw_version(adap);
	if ((adap->flags & MASTER_PF) && state != DEV_STATE_INIT) {
		if (ret == -EINVAL || ret > 0) {
			if (upgrade_fw(adap) >= 0) {
				/*
				 * Note that the chip was reset as part of the
				 * firmware upgrade so we don't reset it again
				 * below and grab the new firmware version.
				 */
				reset = 0;
				ret = t4_check_fw_version(adap);
			}
		}
		if (ret < 0)
			return ret;
	}

	/*
	 * Grab VPD parameters.  This should be done after we establish a
	 * connection to the firmware since some of the VPD parameters
	 * (notably the Core Clock frequency) are retrieved via requests to
	 * the firmware.  On the other hand, we need these fairly early on
	 * so we do this right after getting ahold of the firmware.
	 */
	ret = get_vpd_params(adap, &adap->params.vpd);
	if (ret < 0)
		goto bye;

	/*
	 * Find out what ports are available to us.  Note that we need to do
	 * this before calling adap_init0_no_config() since it needs nports
	 * and portvec ...
	 */
	v =
	    FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) |
	    FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_PORTVEC);
	ret = t4_query_params(adap, adap->mbox, adap->fn, 0, 1, &v, &port_vec);
	if (ret < 0)
		goto bye;

	adap->params.nports = hweight32(port_vec);
	adap->params.portvec = port_vec;

	/*
	 * If the firmware is initialized already (and we're not forcing a
	 * master initialization), note that we're living with existing
	 * adapter parameters.  Otherwise, it's time to try initializing the
	 * adapter ...
	 */
	if (state == DEV_STATE_INIT) {
		dev_info(adap->pdev_dev, "Coming up as %s: "\
			 "Adapter already initialized\n",
			 adap->flags & MASTER_PF ? "MASTER" : "SLAVE");
		adap->flags |= USING_SOFT_PARAMS;
	} else {
		dev_info(adap->pdev_dev, "Coming up as MASTER: "\
			 "Initializing adapter\n");

		/*
		 * If the firmware doesn't support Configuration
		 * Files warn user and exit,
		 */
		if (ret < 0)
			dev_warn(adap->pdev_dev, "Firmware doesn't support "
				 "configuration file.\n");
		if (force_old_init)
			ret = adap_init0_no_config(adap, reset);
		else {
			/*
			 * Find out whether we're dealing with a version of
			 * the firmware which has configuration file support.
			 */
			params[0] = (FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) |
				     FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_CF));
			ret = t4_query_params(adap, adap->mbox, adap->fn, 0, 1,
					      params, val);

			/*
			 * If the firmware doesn't support Configuration
			 * Files, use the old Driver-based, hard-wired
			 * initialization.  Otherwise, try using the
			 * Configuration File support and fall back to the
			 * Driver-based initialization if there's no
			 * Configuration File found.
			 */
			if (ret < 0)
				ret = adap_init0_no_config(adap, reset);
			else {
				/*
				 * The firmware provides us with a memory
				 * buffer where we can load a Configuration
				 * File from the host if we want to override
				 * the Configuration File in flash.
				 */

				ret = adap_init0_config(adap, reset);
				if (ret == -ENOENT) {
					dev_info(adap->pdev_dev,
					    "No Configuration File present "
					    "on adapter.  Using hard-wired "
					    "configuration parameters.\n");
					ret = adap_init0_no_config(adap, reset);
				}
			}
		}
		if (ret < 0) {
			dev_err(adap->pdev_dev,
				"could not initialize adapter, error %d\n",
				-ret);
			goto bye;
		}
	}

	/*
	 * If we're living with non-hard-coded parameters (either from a
	 * Firmware Configuration File or values programmed by a different PF
	 * Driver), give the SGE code a chance to pull in anything that it
	 * needs ...  Note that this must be called after we retrieve our VPD
	 * parameters in order to know how to convert core ticks to seconds.
	 */
	if (adap->flags & USING_SOFT_PARAMS) {
		ret = t4_sge_init(adap);
		if (ret < 0)
			goto bye;
	}

	if (is_bypass_device(adap->pdev->device))
		adap->params.bypass = 1;

	/*
	 * Grab some of our basic fundamental operating parameters.
	 */
#define FW_PARAM_DEV(param) \
	(FW_PARAMS_MNEM(FW_PARAMS_MNEM_DEV) | \
	FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_DEV_##param))

#define FW_PARAM_PFVF(param) \
	FW_PARAMS_MNEM(FW_PARAMS_MNEM_PFVF) | \
	FW_PARAMS_PARAM_X(FW_PARAMS_PARAM_PFVF_##param)|  \
	FW_PARAMS_PARAM_Y(0) | \
	FW_PARAMS_PARAM_Z(0)

	params[0] = FW_PARAM_PFVF(EQ_START);
	params[1] = FW_PARAM_PFVF(L2T_START);
	params[2] = FW_PARAM_PFVF(L2T_END);
	params[3] = FW_PARAM_PFVF(FILTER_START);
	params[4] = FW_PARAM_PFVF(FILTER_END);
	params[5] = FW_PARAM_PFVF(IQFLINT_START);
	ret = t4_query_params(adap, adap->mbox, adap->fn, 0, 6, params, val);
	if (ret < 0)
		goto bye;
	adap->sge.egr_start = val[0];
	adap->l2t_start = val[1];
	adap->l2t_end = val[2];
	adap->tids.ftid_base = val[3];
	adap->tids.nftids = val[4] - val[3] + 1;
	adap->sge.ingr_start = val[5];

	/* query params related to active filter region */
	params[0] = FW_PARAM_PFVF(ACTIVE_FILTER_START);
	params[1] = FW_PARAM_PFVF(ACTIVE_FILTER_END);
	ret = t4_query_params(adap, adap->mbox, adap->fn, 0, 2, params, val);
	/* If Active filter size is set we enable establishing
	 * offload connection through firmware work request
	 */
	if ((val[0] != val[1]) && (ret >= 0)) {
		adap->flags |= FW_OFLD_CONN;
		adap->tids.aftid_base = val[0];
		adap->tids.aftid_end = val[1];
	}

	/*
	 * Get device capabilities so we can determine what resources we need
	 * to manage.
	 */
	memset(&caps_cmd, 0, sizeof(caps_cmd));
	caps_cmd.op_to_write = htonl(FW_CMD_OP(FW_CAPS_CONFIG_CMD) |
				     FW_CMD_REQUEST | FW_CMD_READ);
	caps_cmd.cfvalid_to_len16 = htonl(FW_LEN16(caps_cmd));
	ret = t4_wr_mbox(adap, adap->mbox, &caps_cmd, sizeof(caps_cmd),
			 &caps_cmd);
	if (ret < 0)
		goto bye;

	if (caps_cmd.ofldcaps) {
		/* query offload-related parameters */
		params[0] = FW_PARAM_DEV(NTID);
		params[1] = FW_PARAM_PFVF(SERVER_START);
		params[2] = FW_PARAM_PFVF(SERVER_END);
		params[3] = FW_PARAM_PFVF(TDDP_START);
		params[4] = FW_PARAM_PFVF(TDDP_END);
		params[5] = FW_PARAM_DEV(FLOWC_BUFFIFO_SZ);
		ret = t4_query_params(adap, adap->mbox, adap->fn, 0, 6,
				      params, val);
		if (ret < 0)
			goto bye;
		adap->tids.ntids = val[0];
		adap->tids.natids = min(adap->tids.ntids / 2, MAX_ATIDS);
		adap->tids.stid_base = val[1];
		adap->tids.nstids = val[2] - val[1] + 1;
		/*
		 * Setup server filter region. Divide the availble filter
		 * region into two parts. Regular filters get 1/3rd and server
		 * filters get 2/3rd part. This is only enabled if workarond
		 * path is enabled.
		 * 1. For regular filters.
		 * 2. Server filter: This are special filters which are used
		 * to redirect SYN packets to offload queue.
		 */
		if (adap->flags & FW_OFLD_CONN && !is_bypass(adap)) {
			adap->tids.sftid_base = adap->tids.ftid_base +
					DIV_ROUND_UP(adap->tids.nftids, 3);
			adap->tids.nsftids = adap->tids.nftids -
					 DIV_ROUND_UP(adap->tids.nftids, 3);
			adap->tids.nftids = adap->tids.sftid_base -
						adap->tids.ftid_base;
		}
		adap->vres.ddp.start = val[3];
		adap->vres.ddp.size = val[4] - val[3] + 1;
		adap->params.ofldq_wr_cred = val[5];

		adap->params.offload = 1;
	}
	if (caps_cmd.rdmacaps) {
		params[0] = FW_PARAM_PFVF(STAG_START);
		params[1] = FW_PARAM_PFVF(STAG_END);
		params[2] = FW_PARAM_PFVF(RQ_START);
		params[3] = FW_PARAM_PFVF(RQ_END);
		params[4] = FW_PARAM_PFVF(PBL_START);
		params[5] = FW_PARAM_PFVF(PBL_END);
		ret = t4_query_params(adap, adap->mbox, adap->fn, 0, 6,
				      params, val);
		if (ret < 0)
			goto bye;
		adap->vres.stag.start = val[0];
		adap->vres.stag.size = val[1] - val[0] + 1;
		adap->vres.rq.start = val[2];
		adap->vres.rq.size = val[3] - val[2] + 1;
		adap->vres.pbl.start = val[4];
		adap->vres.pbl.size = val[5] - val[4] + 1;

		params[0] = FW_PARAM_PFVF(SQRQ_START);
		params[1] = FW_PARAM_PFVF(SQRQ_END);
		params[2] = FW_PARAM_PFVF(CQ_START);
		params[3] = FW_PARAM_PFVF(CQ_END);
		params[4] = FW_PARAM_PFVF(OCQ_START);
		params[5] = FW_PARAM_PFVF(OCQ_END);
		ret = t4_query_params(adap, 0, 0, 0, 6, params, val);
		if (ret < 0)
			goto bye;
		adap->vres.qp.start = val[0];
		adap->vres.qp.size = val[1] - val[0] + 1;
		adap->vres.cq.start = val[2];
		adap->vres.cq.size = val[3] - val[2] + 1;
		adap->vres.ocq.start = val[4];
		adap->vres.ocq.size = val[5] - val[4] + 1;
	}
	if (caps_cmd.iscsicaps) {
		params[0] = FW_PARAM_PFVF(ISCSI_START);
		params[1] = FW_PARAM_PFVF(ISCSI_END);
		ret = t4_query_params(adap, adap->mbox, adap->fn, 0, 2,
				      params, val);
		if (ret < 0)
			goto bye;
		adap->vres.iscsi.start = val[0];
		adap->vres.iscsi.size = val[1] - val[0] + 1;
	}
#undef FW_PARAM_PFVF
#undef FW_PARAM_DEV

	/*
	 * These are finalized by FW initialization, load their values now.
	 */
	v = t4_read_reg(adap, TP_TIMER_RESOLUTION);
	adap->params.tp.tre = TIMERRESOLUTION_GET(v);
	adap->params.tp.dack_re = DELAYEDACKRESOLUTION_GET(v);
	t4_read_mtu_tbl(adap, adap->params.mtus, NULL);
	t4_load_mtus(adap, adap->params.mtus, adap->params.a_wnd,
		     adap->params.b_wnd);

	/* MODQ_REQ_MAP defaults to setting queues 0-3 to chan 0-3 */
	for (j = 0; j < NCHAN; j++)
		adap->params.tp.tx_modq[j] = j;

	adap->flags |= FW_OK;
	return 0;

	/*
	 * Something bad happened.  If a command timed out or failed with EIO
	 * FW does not operate within its spec or something catastrophic
	 * happened to HW/FW, stop issuing commands.
	 */
bye:
	if (ret != -ETIMEDOUT && ret != -EIO)
		t4_fw_bye(adap, adap->mbox);
	return ret;
}

/* EEH callbacks */

static pci_ers_result_t eeh_err_detected(struct pci_dev *pdev,
					 pci_channel_state_t state)
{
	int i;
	struct adapter *adap = pci_get_drvdata(pdev);

	if (!adap)
		goto out;

	rtnl_lock();
	adap->flags &= ~FW_OK;
	notify_ulds(adap, CXGB4_STATE_START_RECOVERY);
	for_each_port(adap, i) {
		struct net_device *dev = adap->port[i];

		netif_device_detach(dev);
		netif_carrier_off(dev);
	}
	if (adap->flags & FULL_INIT_DONE)
		cxgb_down(adap);
	rtnl_unlock();
	pci_disable_device(pdev);
out:	return state == pci_channel_io_perm_failure ?
		PCI_ERS_RESULT_DISCONNECT : PCI_ERS_RESULT_NEED_RESET;
}

static pci_ers_result_t eeh_slot_reset(struct pci_dev *pdev)
{
	int i, ret;
	struct fw_caps_config_cmd c;
	struct adapter *adap = pci_get_drvdata(pdev);

	if (!adap) {
		pci_restore_state(pdev);
		pci_save_state(pdev);
		return PCI_ERS_RESULT_RECOVERED;
	}

	if (pci_enable_device(pdev)) {
		dev_err(&pdev->dev, "cannot reenable PCI device after reset\n");
		return PCI_ERS_RESULT_DISCONNECT;
	}

	pci_set_master(pdev);
	pci_restore_state(pdev);
	pci_save_state(pdev);
	pci_cleanup_aer_uncorrect_error_status(pdev);

	if (t4_wait_dev_ready(adap) < 0)
		return PCI_ERS_RESULT_DISCONNECT;
	if (t4_fw_hello(adap, adap->fn, adap->fn, MASTER_MUST, NULL))
		return PCI_ERS_RESULT_DISCONNECT;
	adap->flags |= FW_OK;
	if (adap_init1(adap, &c))
		return PCI_ERS_RESULT_DISCONNECT;

	for_each_port(adap, i) {
		struct port_info *p = adap2pinfo(adap, i);

		ret = t4_alloc_vi(adap, adap->fn, p->tx_chan, adap->fn, 0, 1,
				  NULL, NULL);
		if (ret < 0)
			return PCI_ERS_RESULT_DISCONNECT;
		p->viid = ret;
		p->xact_addr_filt = -1;
	}

	t4_load_mtus(adap, adap->params.mtus, adap->params.a_wnd,
		     adap->params.b_wnd);
	setup_memwin(adap);
	if (cxgb_up(adap))
		return PCI_ERS_RESULT_DISCONNECT;
	return PCI_ERS_RESULT_RECOVERED;
}

static void eeh_resume(struct pci_dev *pdev)
{
	int i;
	struct adapter *adap = pci_get_drvdata(pdev);

	if (!adap)
		return;

	rtnl_lock();
	for_each_port(adap, i) {
		struct net_device *dev = adap->port[i];

		if (netif_running(dev)) {
			link_start(dev);
			cxgb_set_rxmode(dev);
		}
		netif_device_attach(dev);
	}
	rtnl_unlock();
}

static const struct pci_error_handlers cxgb4_eeh = {
	.error_detected = eeh_err_detected,
	.slot_reset     = eeh_slot_reset,
	.resume         = eeh_resume,
};

static inline bool is_10g_port(const struct link_config *lc)
{
	return (lc->supported & FW_PORT_CAP_SPEED_10G) != 0;
}

static inline void init_rspq(struct sge_rspq *q, u8 timer_idx, u8 pkt_cnt_idx,
			     unsigned int size, unsigned int iqe_size)
{
	q->intr_params = QINTR_TIMER_IDX(timer_idx) |
			 (pkt_cnt_idx < SGE_NCOUNTERS ? QINTR_CNT_EN : 0);
	q->pktcnt_idx = pkt_cnt_idx < SGE_NCOUNTERS ? pkt_cnt_idx : 0;
	q->iqe_len = iqe_size;
	q->size = size;
}

/*
 * Perform default configuration of DMA queues depending on the number and type
 * of ports we found and the number of available CPUs.  Most settings can be
 * modified by the admin prior to actual use.
 */
static void cfg_queues(struct adapter *adap)
{
	struct sge *s = &adap->sge;
	int i, q10g = 0, n10g = 0, qidx = 0;

	for_each_port(adap, i)
		n10g += is_10g_port(&adap2pinfo(adap, i)->link_cfg);

	/*
	 * We default to 1 queue per non-10G port and up to # of cores queues
	 * per 10G port.
	 */
	if (n10g)
		q10g = (MAX_ETH_QSETS - (adap->params.nports - n10g)) / n10g;
	if (q10g > netif_get_num_default_rss_queues())
		q10g = netif_get_num_default_rss_queues();

	for_each_port(adap, i) {
		struct port_info *pi = adap2pinfo(adap, i);

		pi->first_qset = qidx;
		pi->nqsets = is_10g_port(&pi->link_cfg) ? q10g : 1;
		qidx += pi->nqsets;
	}

	s->ethqsets = qidx;
	s->max_ethqsets = qidx;   /* MSI-X may lower it later */

	if (is_offload(adap)) {
		/*
		 * For offload we use 1 queue/channel if all ports are up to 1G,
		 * otherwise we divide all available queues amongst the channels
		 * capped by the number of available cores.
		 */
		if (n10g) {
			i = min_t(int, ARRAY_SIZE(s->ofldrxq),
				  num_online_cpus());
			s->ofldqsets = roundup(i, adap->params.nports);
		} else
			s->ofldqsets = adap->params.nports;
		/* For RDMA one Rx queue per channel suffices */
		s->rdmaqs = adap->params.nports;
	}

	for (i = 0; i < ARRAY_SIZE(s->ethrxq); i++) {
		struct sge_eth_rxq *r = &s->ethrxq[i];

		init_rspq(&r->rspq, 0, 0, 1024, 64);
		r->fl.size = 72;
	}

	for (i = 0; i < ARRAY_SIZE(s->ethtxq); i++)
		s->ethtxq[i].q.size = 1024;

	for (i = 0; i < ARRAY_SIZE(s->ctrlq); i++)
		s->ctrlq[i].q.size = 512;

	for (i = 0; i < ARRAY_SIZE(s->ofldtxq); i++)
		s->ofldtxq[i].q.size = 1024;

	for (i = 0; i < ARRAY_SIZE(s->ofldrxq); i++) {
		struct sge_ofld_rxq *r = &s->ofldrxq[i];

		init_rspq(&r->rspq, 0, 0, 1024, 64);
		r->rspq.uld = CXGB4_ULD_ISCSI;
		r->fl.size = 72;
	}

	for (i = 0; i < ARRAY_SIZE(s->rdmarxq); i++) {
		struct sge_ofld_rxq *r = &s->rdmarxq[i];

		init_rspq(&r->rspq, 0, 0, 511, 64);
		r->rspq.uld = CXGB4_ULD_RDMA;
		r->fl.size = 72;
	}

	init_rspq(&s->fw_evtq, 6, 0, 512, 64);
	init_rspq(&s->intrq, 6, 0, 2 * MAX_INGQ, 64);
}

/*
 * Reduce the number of Ethernet queues across all ports to at most n.
 * n provides at least one queue per port.
 */
static void reduce_ethqs(struct adapter *adap, int n)
{
	int i;
	struct port_info *pi;

	while (n < adap->sge.ethqsets)
		for_each_port(adap, i) {
			pi = adap2pinfo(adap, i);
			if (pi->nqsets > 1) {
				pi->nqsets--;
				adap->sge.ethqsets--;
				if (adap->sge.ethqsets <= n)
					break;
			}
		}

	n = 0;
	for_each_port(adap, i) {
		pi = adap2pinfo(adap, i);
		pi->first_qset = n;
		n += pi->nqsets;
	}
}

/* 2 MSI-X vectors needed for the FW queue and non-data interrupts */
#define EXTRA_VECS 2

static int enable_msix(struct adapter *adap)
{
	int ofld_need = 0;
	int i, err, want, need;
	struct sge *s = &adap->sge;
	unsigned int nchan = adap->params.nports;
	struct msix_entry entries[MAX_INGQ + 1];

	for (i = 0; i < ARRAY_SIZE(entries); ++i)
		entries[i].entry = i;

	want = s->max_ethqsets + EXTRA_VECS;
	if (is_offload(adap)) {
		want += s->rdmaqs + s->ofldqsets;
		/* need nchan for each possible ULD */
		ofld_need = 2 * nchan;
	}
	need = adap->params.nports + EXTRA_VECS + ofld_need;

	while ((err = pci_enable_msix(adap->pdev, entries, want)) >= need)
		want = err;

	if (!err) {
		/*
		 * Distribute available vectors to the various queue groups.
		 * Every group gets its minimum requirement and NIC gets top
		 * priority for leftovers.
		 */
		i = want - EXTRA_VECS - ofld_need;
		if (i < s->max_ethqsets) {
			s->max_ethqsets = i;
			if (i < s->ethqsets)
				reduce_ethqs(adap, i);
		}
		if (is_offload(adap)) {
			i = want - EXTRA_VECS - s->max_ethqsets;
			i -= ofld_need - nchan;
			s->ofldqsets = (i / nchan) * nchan;  /* round down */
		}
		for (i = 0; i < want; ++i)
			adap->msix_info[i].vec = entries[i].vector;
	} else if (err > 0)
		dev_info(adap->pdev_dev,
			 "only %d MSI-X vectors left, not using MSI-X\n", err);
	return err;
}

#undef EXTRA_VECS

static int init_rss(struct adapter *adap)
{
	unsigned int i, j;

	for_each_port(adap, i) {
		struct port_info *pi = adap2pinfo(adap, i);

		pi->rss = kcalloc(pi->rss_size, sizeof(u16), GFP_KERNEL);
		if (!pi->rss)
			return -ENOMEM;
		for (j = 0; j < pi->rss_size; j++)
			pi->rss[j] = ethtool_rxfh_indir_default(j, pi->nqsets);
	}
	return 0;
}

static void print_port_info(const struct net_device *dev)
{
	static const char *base[] = {
		"R XFI", "R XAUI", "T SGMII", "T XFI", "T XAUI", "KX4", "CX4",
		"KX", "KR", "R SFP+", "KR/KX", "KR/KX/KX4"
	};

	char buf[80];
	char *bufp = buf;
	const char *spd = "";
	const struct port_info *pi = netdev_priv(dev);
	const struct adapter *adap = pi->adapter;

	if (adap->params.pci.speed == PCI_EXP_LNKSTA_CLS_2_5GB)
		spd = " 2.5 GT/s";
	else if (adap->params.pci.speed == PCI_EXP_LNKSTA_CLS_5_0GB)
		spd = " 5 GT/s";

	if (pi->link_cfg.supported & FW_PORT_CAP_SPEED_100M)
		bufp += sprintf(bufp, "100/");
	if (pi->link_cfg.supported & FW_PORT_CAP_SPEED_1G)
		bufp += sprintf(bufp, "1000/");
	if (pi->link_cfg.supported & FW_PORT_CAP_SPEED_10G)
		bufp += sprintf(bufp, "10G/");
	if (bufp != buf)
		--bufp;
	sprintf(bufp, "BASE-%s", base[pi->port_type]);

	netdev_info(dev, "Chelsio %s rev %d %s %sNIC PCIe x%d%s%s\n",
		    adap->params.vpd.id, adap->params.rev, buf,
		    is_offload(adap) ? "R" : "", adap->params.pci.width, spd,
		    (adap->flags & USING_MSIX) ? " MSI-X" :
		    (adap->flags & USING_MSI) ? " MSI" : "");
	netdev_info(dev, "S/N: %s, E/C: %s\n",
		    adap->params.vpd.sn, adap->params.vpd.ec);
}

static void enable_pcie_relaxed_ordering(struct pci_dev *dev)
{
	pcie_capability_set_word(dev, PCI_EXP_DEVCTL, PCI_EXP_DEVCTL_RELAX_EN);
}

/*
 * Free the following resources:
 * - memory used for tables
 * - MSI/MSI-X
 * - net devices
 * - resources FW is holding for us
 */
static void free_some_resources(struct adapter *adapter)
{
	unsigned int i;

	t4_free_mem(adapter->l2t);
	t4_free_mem(adapter->tids.tid_tab);
	disable_msi(adapter);

	for_each_port(adapter, i)
		if (adapter->port[i]) {
			kfree(adap2pinfo(adapter, i)->rss);
			free_netdev(adapter->port[i]);
		}
	if (adapter->flags & FW_OK)
		t4_fw_bye(adapter, adapter->fn);
}

#define TSO_FLAGS (NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_TSO_ECN)
#define VLAN_FEAT (NETIF_F_SG | NETIF_F_IP_CSUM | TSO_FLAGS | \
		   NETIF_F_IPV6_CSUM | NETIF_F_HIGHDMA)

static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
{
	int func, i, err;
	struct port_info *pi;
	bool highdma = false;
	struct adapter *adapter = NULL;

	printk_once(KERN_INFO "%s - version %s\n", DRV_DESC, DRV_VERSION);

	err = pci_request_regions(pdev, KBUILD_MODNAME);
	if (err) {
		/* Just info, some other driver may have claimed the device. */
		dev_info(&pdev->dev, "cannot obtain PCI resources\n");
		return err;
	}

	/* We control everything through one PF */
	func = PCI_FUNC(pdev->devfn);
	if (func != ent->driver_data) {
		pci_save_state(pdev);        /* to restore SR-IOV later */
		goto sriov;
	}

	err = pci_enable_device(pdev);
	if (err) {
		dev_err(&pdev->dev, "cannot enable PCI device\n");
		goto out_release_regions;
	}

	if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(64))) {
		highdma = true;
		err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64));
		if (err) {
			dev_err(&pdev->dev, "unable to obtain 64-bit DMA for "
				"coherent allocations\n");
			goto out_disable_device;
		}
	} else {
		err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
		if (err) {
			dev_err(&pdev->dev, "no usable DMA configuration\n");
			goto out_disable_device;
		}
	}

	pci_enable_pcie_error_reporting(pdev);
	enable_pcie_relaxed_ordering(pdev);
	pci_set_master(pdev);
	pci_save_state(pdev);

	adapter = kzalloc(sizeof(*adapter), GFP_KERNEL);
	if (!adapter) {
		err = -ENOMEM;
		goto out_disable_device;
	}

	adapter->regs = pci_ioremap_bar(pdev, 0);
	if (!adapter->regs) {
		dev_err(&pdev->dev, "cannot map device registers\n");
		err = -ENOMEM;
		goto out_free_adapter;
	}

	adapter->pdev = pdev;
	adapter->pdev_dev = &pdev->dev;
	adapter->mbox = func;
	adapter->fn = func;
	adapter->msg_enable = dflt_msg_enable;
	memset(adapter->chan_map, 0xff, sizeof(adapter->chan_map));

	spin_lock_init(&adapter->stats_lock);
	spin_lock_init(&adapter->tid_release_lock);

	INIT_WORK(&adapter->tid_release_task, process_tid_release_list);
	INIT_WORK(&adapter->db_full_task, process_db_full);
	INIT_WORK(&adapter->db_drop_task, process_db_drop);

	err = t4_prep_adapter(adapter);
	if (err)
		goto out_unmap_bar;
	setup_memwin(adapter);
	err = adap_init0(adapter);
	setup_memwin_rdma(adapter);
	if (err)
		goto out_unmap_bar;

	for_each_port(adapter, i) {
		struct net_device *netdev;

		netdev = alloc_etherdev_mq(sizeof(struct port_info),
					   MAX_ETH_QSETS);
		if (!netdev) {
			err = -ENOMEM;
			goto out_free_dev;
		}

		SET_NETDEV_DEV(netdev, &pdev->dev);

		adapter->port[i] = netdev;
		pi = netdev_priv(netdev);
		pi->adapter = adapter;
		pi->xact_addr_filt = -1;
		pi->port_id = i;
		netdev->irq = pdev->irq;

		netdev->hw_features = NETIF_F_SG | TSO_FLAGS |
			NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
			NETIF_F_RXCSUM | NETIF_F_RXHASH |
			NETIF_F_HW_VLAN_TX | NETIF_F_HW_VLAN_RX;
		if (highdma)
			netdev->hw_features |= NETIF_F_HIGHDMA;
		netdev->features |= netdev->hw_features;
		netdev->vlan_features = netdev->features & VLAN_FEAT;

		netdev->priv_flags |= IFF_UNICAST_FLT;

		netdev->netdev_ops = &cxgb4_netdev_ops;
		SET_ETHTOOL_OPS(netdev, &cxgb_ethtool_ops);
	}

	pci_set_drvdata(pdev, adapter);

	if (adapter->flags & FW_OK) {
		err = t4_port_init(adapter, func, func, 0);
		if (err)
			goto out_free_dev;
	}

	/*
	 * Configure queues and allocate tables now, they can be needed as
	 * soon as the first register_netdev completes.
	 */
	cfg_queues(adapter);

	adapter->l2t = t4_init_l2t();
	if (!adapter->l2t) {
		/* We tolerate a lack of L2T, giving up some functionality */
		dev_warn(&pdev->dev, "could not allocate L2T, continuing\n");
		adapter->params.offload = 0;
	}

	if (is_offload(adapter) && tid_init(&adapter->tids) < 0) {
		dev_warn(&pdev->dev, "could not allocate TID table, "
			 "continuing\n");
		adapter->params.offload = 0;
	}

	/* See what interrupts we'll be using */
	if (msi > 1 && enable_msix(adapter) == 0)
		adapter->flags |= USING_MSIX;
	else if (msi > 0 && pci_enable_msi(pdev) == 0)
		adapter->flags |= USING_MSI;

	err = init_rss(adapter);
	if (err)
		goto out_free_dev;

	/*
	 * The card is now ready to go.  If any errors occur during device
	 * registration we do not fail the whole card but rather proceed only
	 * with the ports we manage to register successfully.  However we must
	 * register at least one net device.
	 */
	for_each_port(adapter, i) {
		pi = adap2pinfo(adapter, i);
		netif_set_real_num_tx_queues(adapter->port[i], pi->nqsets);
		netif_set_real_num_rx_queues(adapter->port[i], pi->nqsets);

		err = register_netdev(adapter->port[i]);
		if (err)
			break;
		adapter->chan_map[pi->tx_chan] = i;
		print_port_info(adapter->port[i]);
	}
	if (i == 0) {
		dev_err(&pdev->dev, "could not register any net devices\n");
		goto out_free_dev;
	}
	if (err) {
		dev_warn(&pdev->dev, "only %d net devices registered\n", i);
		err = 0;
	}

	if (cxgb4_debugfs_root) {
		adapter->debugfs_root = debugfs_create_dir(pci_name(pdev),
							   cxgb4_debugfs_root);
		setup_debugfs(adapter);
	}

	/* PCIe EEH recovery on powerpc platforms needs fundamental reset */
	pdev->needs_freset = 1;

	if (is_offload(adapter))
		attach_ulds(adapter);

sriov:
#ifdef CONFIG_PCI_IOV
	if (func < ARRAY_SIZE(num_vf) && num_vf[func] > 0)
		if (pci_enable_sriov(pdev, num_vf[func]) == 0)
			dev_info(&pdev->dev,
				 "instantiated %u virtual functions\n",
				 num_vf[func]);
#endif
	return 0;

 out_free_dev:
	free_some_resources(adapter);
 out_unmap_bar:
	iounmap(adapter->regs);
 out_free_adapter:
	kfree(adapter);
 out_disable_device:
	pci_disable_pcie_error_reporting(pdev);
	pci_disable_device(pdev);
 out_release_regions:
	pci_release_regions(pdev);
	pci_set_drvdata(pdev, NULL);
	return err;
}

static void remove_one(struct pci_dev *pdev)
{
	struct adapter *adapter = pci_get_drvdata(pdev);

#ifdef CONFIG_PCI_IOV
	pci_disable_sriov(pdev);

#endif

	if (adapter) {
		int i;

		if (is_offload(adapter))
			detach_ulds(adapter);

		for_each_port(adapter, i)
			if (adapter->port[i]->reg_state == NETREG_REGISTERED)
				unregister_netdev(adapter->port[i]);

		if (adapter->debugfs_root)
			debugfs_remove_recursive(adapter->debugfs_root);

		/* If we allocated filters, free up state associated with any
		 * valid filters ...
		 */
		if (adapter->tids.ftid_tab) {
			struct filter_entry *f = &adapter->tids.ftid_tab[0];
			for (i = 0; i < adapter->tids.nftids; i++, f++)
				if (f->valid)
					clear_filter(adapter, f);
		}

		if (adapter->flags & FULL_INIT_DONE)
			cxgb_down(adapter);

		free_some_resources(adapter);
		iounmap(adapter->regs);
		kfree(adapter);
		pci_disable_pcie_error_reporting(pdev);
		pci_disable_device(pdev);
		pci_release_regions(pdev);
		pci_set_drvdata(pdev, NULL);
	} else
		pci_release_regions(pdev);
}

static struct pci_driver cxgb4_driver = {
	.name     = KBUILD_MODNAME,
	.id_table = cxgb4_pci_tbl,
	.probe    = init_one,
	.remove   = remove_one,
	.err_handler = &cxgb4_eeh,
};

static int __init cxgb4_init_module(void)
{
	int ret;

	workq = create_singlethread_workqueue("cxgb4");
	if (!workq)
		return -ENOMEM;

	/* Debugfs support is optional, just warn if this fails */
	cxgb4_debugfs_root = debugfs_create_dir(KBUILD_MODNAME, NULL);
	if (!cxgb4_debugfs_root)
		pr_warning("could not create debugfs entry, continuing\n");

	ret = pci_register_driver(&cxgb4_driver);
	if (ret < 0)
		debugfs_remove(cxgb4_debugfs_root);
	return ret;
}

static void __exit cxgb4_cleanup_module(void)
{
	pci_unregister_driver(&cxgb4_driver);
	debugfs_remove(cxgb4_debugfs_root);  /* NULL ok */
	flush_workqueue(workq);
	destroy_workqueue(workq);
}

module_init(cxgb4_init_module);
module_exit(cxgb4_cleanup_module);