path: root/linux/src/drivers/net/starfire.c

/* starfire.c: Linux device driver for the Adaptec Starfire network adapter. */
/*
	Written/Copyright 1998-2003 by Donald Becker.

	This software may be used and distributed according to the terms of
	the GNU General Public License (GPL), incorporated herein by reference.
	Drivers based on or derived from this code fall under the GPL and must
	retain the authorship, copyright and license notice.  This file is not
	a complete program and may only be used when the entire operating
	system is licensed under the GPL.

	The author may be reached as becker@scyld.com, or C/O
	Scyld Computing Corporation
	914 Bay Ridge Road, Suite 220
	Annapolis MD 21403

	Support information and updates available at
	http://www.scyld.com/network/starfire.html
*/

/* These identify the driver base version and may not be removed. */
static const char version1[] =
"starfire.c:v1.09 7/22/2003  Copyright by Donald Becker <becker@scyld.com>\n";
static const char version2[] =
" Updates and info at http://www.scyld.com/network/starfire.html\n";

/* The user-configurable values.
   These may be modified when a driver module is loaded.*/

/* Used for tuning interrupt latency vs. overhead. */
static int interrupt_mitigation = 0x0;

/* Message enable level: 0..31 = no..all messages.  See NETIF_MSG docs. */
static int debug = 2;

/* Maximum events (Rx packets, etc.) to handle at each interrupt. */
static int max_interrupt_work = 20;

/* Maximum number of multicast addresses to filter (vs. rx-all-multicast).
   The Starfire has a 512 element hash table based on the Ethernet CRC.  */
static int multicast_filter_limit = 32;

/* Set the copy breakpoint for the copy-only-tiny-frames scheme.
   Setting to > 1518 effectively disables this feature. */
static int rx_copybreak = 0;

/* Used to pass the media type, etc.
   Both 'options[]' and 'full_duplex[]' exist for driver interoperability,
   however full_duplex[] should never be used in new configurations.
   The media type is usually passed in 'options[]'.
    The default is autonegotation for speed and duplex.
	This should rarely be overridden.
    Use option values 0x10/0x20 for 10Mbps, 0x100,0x200 for 100Mbps.
    Use option values 0x10 and 0x100 for forcing half duplex fixed speed.
    Use option values 0x20 and 0x200 for forcing full duplex operation.
*/
#define MAX_UNITS 8		/* More are supported, limit only on options */
static int options[MAX_UNITS] = {-1, -1, -1, -1, -1, -1, -1, -1};
static int full_duplex[MAX_UNITS] = {-1, -1, -1, -1, -1, -1, -1, -1};

/* Automatically extracted configuration info:
probe-func: starfire_probe
config-in: tristate 'Adaptec DuraLAN ("starfire") series PCI Ethernet support' CONFIG_DURLAN

c-help-name: Adaptec DuraLAN ("starfire") series PCI Ethernet support
c-help-symbol: CONFIG_DURALAN
c-help: This driver is for the Adaptec DuraLAN series, the 6915, 62022
c-help: and 62044 boards.
c-help: Design information, usage details and updates are available from
c-help: http://www.scyld.com/network/starfire.html
*/

/* Operational parameters that are set at compile time. */

/* The "native" ring sizes are either 256 or 2048.
   However in some modes a descriptor may be marked to wrap the ring earlier.
   The driver allocates a single page for each descriptor ring, constraining
   the maximum size in an architecture-dependent way.
*/
#define RX_RING_SIZE	256
#define TX_RING_SIZE	32
/* The completion queues are fixed at 1024 entries i.e. 4K or 8KB. */
#define DONE_Q_SIZE	1024

/* Operational parameters that usually are not changed. */
/* Time in jiffies before concluding the transmitter is hung. */
#define TX_TIMEOUT  (6*HZ)

/* Allocation size of Rx buffers with normal sized Ethernet frames.
   Do not change this value without good reason.  This is not a limit,
   but a way to keep a consistent allocation size among drivers.
 */
#define PKT_BUF_SZ		1536

#ifndef __KERNEL__
#define __KERNEL__
#endif
#if !defined(__OPTIMIZE__)
#warning  You must compile this file with the correct options!
#warning  See the last lines of the source file.
#error You must compile this driver with "-O".
#endif

/* Include files, designed to support most kernel versions 2.0.0 and later. */
#include <linux/config.h>
#if defined(CONFIG_SMP) && ! defined(__SMP__)
#define __SMP__
#endif
#if defined(MODULE) && defined(CONFIG_MODVERSIONS) && ! defined(MODVERSIONS)
#define MODVERSIONS
#endif

#include <linux/version.h>
#if defined(MODVERSIONS)
#include <linux/modversions.h>
#endif
#include <linux/module.h>

#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/timer.h>
#include <linux/errno.h>
#include <linux/ioport.h>
#if LINUX_VERSION_CODE >= 0x20400
#include <linux/slab.h>
#else
#include <linux/malloc.h>
#endif
#include <linux/interrupt.h>
#include <linux/pci.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/skbuff.h>
#include <asm/processor.h>		/* Processor type for cache alignment. */
#include <asm/bitops.h>
#include <asm/io.h>

#ifdef INLINE_PCISCAN
#include "k_compat.h"
#else
#include "pci-scan.h"
#include "kern_compat.h"
#endif

/* Condensed operations for readability.
   Compatibility defines are in kern_compat.h */

#define virt_to_le32desc(addr)  cpu_to_le32(virt_to_bus(addr))
#define le32desc_to_virt(addr)  bus_to_virt(le32_to_cpu(addr))

#if (LINUX_VERSION_CODE >= 0x20100)  &&  defined(MODULE)
char kernel_version[] = UTS_RELEASE;
#endif

MODULE_AUTHOR("Donald Becker <becker@scyld.com>");
MODULE_DESCRIPTION("Adaptec Starfire Ethernet driver");
MODULE_LICENSE("GPL");
MODULE_PARM(debug, "i");
MODULE_PARM(options, "1-" __MODULE_STRING(MAX_UNITS) "i");
MODULE_PARM(rx_copybreak, "i");
MODULE_PARM(full_duplex, "1-" __MODULE_STRING(MAX_UNITS) "i");
MODULE_PARM(multicast_filter_limit, "i");
MODULE_PARM(max_interrupt_work, "i");
MODULE_PARM_DESC(debug, "Driver message enable level (0-31)");
MODULE_PARM_DESC(options, "Force transceiver type or fixed speed+duplex");
MODULE_PARM_DESC(max_interrupt_work,
				 "Driver maximum events handled per interrupt");
MODULE_PARM_DESC(full_duplex,
				 "Non-zero to set forced full duplex (deprecated).");
MODULE_PARM_DESC(rx_copybreak,
				 "Breakpoint in bytes for copy-only-tiny-frames");
MODULE_PARM_DESC(multicast_filter_limit,
				 "Multicast addresses before switching to Rx-all-multicast");

/*
				Theory of Operation

I. Board Compatibility

This driver is for the Adaptec 6915 DuraLAN "Starfire" 64 bit PCI Ethernet
adapter, and the multiport boards using the same chip.

II. Board-specific settings

III. Driver operation

IIIa. Ring buffers

The Starfire hardware uses multiple fixed-size descriptor queues/rings.  The
ring sizes are set fixed by the hardware, but may optionally be wrapped
earlier by the END bit in the descriptor.
This driver uses that hardware queue size for the Rx ring, where a large
number of entries has no ill effect beyond increases the potential backlog.
The Tx ring is wrapped with the END bit, since a large hardware Tx queue
disables the queue layer priority ordering and we have no mechanism to
utilize the hardware two-level priority queue.  When modifying the
RX/TX_RING_SIZE pay close attention to page sizes and the ring-empty warning
levels.

IIIb/c. Transmit/Receive Structure

See the Adaptec manual for the many possible structures, and options for
each structure.  There are far too many to document here.

For transmit this driver uses type 1 transmit descriptors, and relies on
automatic minimum-length padding.  It does not use the completion queue
consumer index, but instead checks for non-zero status entries.

For receive this driver uses type 0 receive descriptors.  The driver
allocates full frame size skbuffs for the Rx ring buffers, so all frames
should fit in a single descriptor.  The driver does not use the completion
queue consumer index, but instead checks for non-zero status entries.

When an incoming frame is less than RX_COPYBREAK bytes long, a fresh skbuff
is allocated and the frame is copied to the new skbuff.  When the incoming
frame is larger, the skbuff is passed directly up the protocol stack.
Buffers consumed this way are replaced by newly allocated skbuffs in a later
phase of receive.

A notable aspect of operation is that unaligned buffers are not permitted by
the Starfire hardware.  The IP header at offset 14 in an ethernet frame thus
isn't longword aligned, which may cause problems on some machine
e.g. Alphas.  Copied frames are put into the skbuff at an offset of "+2",
16-byte aligning the IP header.

IIId. Synchronization

The driver runs as two independent, single-threaded flows of control.  One
is the send-packet routine, which enforces single-threaded use by the
dev->tbusy flag.  The other thread is the interrupt handler, which is single
threaded by the hardware and interrupt handling software.

The send packet thread has partial control over the Tx ring and 'dev->tbusy'
flag.  It sets the tbusy flag whenever it's queuing a Tx packet. If the next
queue slot is empty, it clears the tbusy flag when finished otherwise it sets
the 'lp->tx_full' flag.

The interrupt handler has exclusive control over the Rx ring and records stats
from the Tx ring.  After reaping the stats, it marks the Tx queue entry as
empty by incrementing the dirty_tx mark. Iff the 'lp->tx_full' flag is set, it
clears both the tx_full and tbusy flags.

IV. Notes

IVb. References

The Adaptec Starfire manuals, available only from Adaptec.
http://www.scyld.com/expert/100mbps.html
http://www.scyld.com/expert/NWay.html

IVc. Errata

*/



static void *starfire_probe1(struct pci_dev *pdev, void *init_dev,
							 long ioaddr, int irq, int chip_idx, int find_cnt);
static int starfire_pwr_event(void *dev_instance, int event);
enum chip_capability_flags {CanHaveMII=1, };
#define PCI_IOTYPE (PCI_USES_MASTER | PCI_USES_MEM | PCI_ADDR0)
/* And maps in 0.5MB(!) -- no I/O mapping here!  */
#define MEM_ADDR_SZ 0x80000

#if 0 && (defined(__x86_64) || defined(__alpha__))
/* Enable 64 bit address modes. */
#define STARFIRE_ADDR_64BITS 1
#endif

static struct pci_id_info pci_id_tbl[] = {
	{"Adaptec Starfire 6915", { 0x69159004, 0xffffffff, },
	 PCI_IOTYPE, MEM_ADDR_SZ, CanHaveMII},
	{0,},						/* 0 terminated list. */
};

struct drv_id_info starfire_drv_id = {
	"starfire", PCI_HOTSWAP, PCI_CLASS_NETWORK_ETHERNET<<8, pci_id_tbl,
	starfire_probe1, starfire_pwr_event };

/* Offsets to the device registers.
   Unlike software-only systems, device drivers interact with complex hardware.
   It's not useful to define symbolic names for every register bit in the
   device.  The name can only partially document the semantics and make
   the driver longer and more difficult to read.
   In general, only the important configuration values or bits changed
   multiple times should be defined symbolically.
*/
enum register_offsets {
	PCIDeviceConfig=0x50040, GenCtrl=0x50070, IntrTimerCtrl=0x50074,
	IntrClear=0x50080, IntrStatus=0x50084, IntrEnable=0x50088,
	MIICtrl=0x52000, StationAddr=0x50120, EEPROMCtrl=0x51000,
	TxDescCtrl=0x50090,
	TxRingPtr=0x50098, HiPriTxRingPtr=0x50094, /* Low and High priority. */
	TxRingHiAddr=0x5009C,		/* 64 bit address extension. */
	TxProducerIdx=0x500A0, TxConsumerIdx=0x500A4,
	TxThreshold=0x500B0,
	CompletionHiAddr=0x500B4, TxCompletionAddr=0x500B8,
	RxCompletionAddr=0x500BC, RxCompletionQ2Addr=0x500C0,
	CompletionQConsumerIdx=0x500C4,
	RxDescQCtrl=0x500D4, RxDescQHiAddr=0x500DC, RxDescQAddr=0x500E0,
	RxDescQIdx=0x500E8, RxDMAStatus=0x500F0, RxFilterMode=0x500F4,
	TxMode=0x55000,
};

/* Bits in the interrupt status/mask registers. */
enum intr_status_bits {
	IntrNormalSummary=0x8000,	IntrAbnormalSummary=0x02000000,
	IntrRxDone=0x0300, IntrRxEmpty=0x10040, IntrRxPCIErr=0x80000,
	IntrTxDone=0x4000, IntrTxEmpty=0x1000, IntrTxPCIErr=0x80000,
	StatsMax=0x08000000, LinkChange=0xf0000000,
	IntrTxDataLow=0x00040000,
	IntrPCIPin=0x01,
};

/* Bits in the RxFilterMode register. */
enum rx_mode_bits {
	AcceptBroadcast=0x04, AcceptAllMulticast=0x02, AcceptAll=0x01,
	AcceptMulticast=0x10, AcceptMyPhys=0xE040,
};

/* Misc. bits.  Symbolic names so that may be searched for. */
enum misc_bits {
	ChipResetCmd=1,				/* PCIDeviceConfig */
	PCIIntEnb=0x00800000,		/* PCIDeviceConfig */
	TxEnable=0x0A, RxEnable=0x05, SoftIntr=0x100, /* GenCtrl */
};

/* The Rx and Tx buffer descriptors. */
struct starfire_rx_desc {
	u32 rxaddr;					/* Optionally 64 bits. */
#if defined(STARFIRE_ADDR_64BITS)
	u32 rxaddr_hi;					/* Optionally 64 bits. */
#endif
};
enum rx_desc_bits {
	RxDescValid=1, RxDescEndRing=2,
};

/* Completion queue entry.
   You must update the page allocation, init_ring and the shift count in rx()
   if using a larger format. */
struct rx_done_desc {
	u32 status;					/* Low 16 bits is length. */
#ifdef full_rx_status
	u32 status2;
	u16 vlanid;
	u16 csum; 			/* partial checksum */
	u32 timestamp;
#endif
};
enum rx_done_bits {
	RxOK=0x20000000, RxFIFOErr=0x10000000, RxBufQ2=0x08000000,
};

/* Type 1 Tx descriptor. */
struct starfire_tx_desc {
	u32 status;					/* Upper bits are status, lower 16 length. */
	u32 addr;
};
enum tx_desc_bits {
	TxDescID=0xB1010000,		/* Also marks single fragment, add CRC.  */
	TxDescIntr=0x08000000, TxRingWrap=0x04000000,
};
struct tx_done_report {
	u32 status;					/* timestamp, index. */
#if 0
	u32 intrstatus;				/* interrupt status */
#endif
};

#define PRIV_ALIGN	15 	/* Required alignment mask */
struct netdev_private {
	/* Descriptor rings first for alignment. */
	struct starfire_rx_desc *rx_ring;
	struct starfire_tx_desc *tx_ring;
	struct net_device *next_module;		/* Link for devices of this type. */
	void *priv_addr;					/* Unaligned address for kfree */
	const char *product_name;
	/* The addresses of rx/tx-in-place skbuffs. */
	struct sk_buff* rx_skbuff[RX_RING_SIZE];
	struct sk_buff* tx_skbuff[TX_RING_SIZE];
	u8 pad0[100];						/* Impact padding */
	/* Pointers to completion queues (full pages).  Cache line pad.. */
	struct rx_done_desc *rx_done_q  __attribute__((aligned (L1_CACHE_BYTES)));
	unsigned int rx_done;
	struct tx_done_report *tx_done_q __attribute__((aligned (L1_CACHE_BYTES)));
	unsigned int tx_done;

	struct net_device_stats stats;
	struct timer_list timer;	/* Media monitoring timer. */
	int msg_level;
	int chip_id, drv_flags;
	struct pci_dev *pci_dev;
	/* Frequently used values: keep some adjacent for cache effect. */
	int max_interrupt_work;
	int intr_enable;
	unsigned int restore_intr_enable:1;	/* Set if temporarily masked.  */
	unsigned int polling:1;				/* Erk, IRQ err. */

	unsigned int cur_rx, dirty_rx;		/* Producer/consumer ring indices */
	unsigned int rx_buf_sz;				/* Based on MTU+slack. */
	int rx_copybreak;

	unsigned int cur_tx, dirty_tx;
	unsigned int tx_full:1;				/* The Tx queue is full. */
	/* These values keep track of the transceiver/media in use. */
	unsigned int full_duplex:1,			/* Full-duplex operation requested. */
		medialock:1,					/* Xcvr set to fixed speed/duplex. */
		rx_flowctrl:1,
		tx_flowctrl:1;					/* Use 802.3x flow control. */
	unsigned int default_port;			/* Last dev->if_port value. */
	u32 tx_mode;
	u8 tx_threshold;
	u32 cur_rx_mode;
	u16 mc_filter[32];
	int multicast_filter_limit;

	/* MII transceiver section. */
	int mii_cnt;						/* MII device addresses. */
	u16 advertising;					/* NWay media advertisement */
	unsigned char phys[2];				/* MII device addresses. */
};

static int  mdio_read(struct net_device *dev, int phy_id, int location);
static void mdio_write(struct net_device *dev, int phy_id, int location,
					   int value);
static int  netdev_open(struct net_device *dev);
static int  change_mtu(struct net_device *dev, int new_mtu);
static void check_duplex(struct net_device *dev);
static void netdev_timer(unsigned long data);
static void tx_timeout(struct net_device *dev);
static void init_ring(struct net_device *dev);
static int  start_tx(struct sk_buff *skb, struct net_device *dev);
static void intr_handler(int irq, void *dev_instance, struct pt_regs *regs);
static void netdev_error(struct net_device *dev, int intr_status);
static int  netdev_rx(struct net_device *dev);
static void netdev_error(struct net_device *dev, int intr_status);
static void set_rx_mode(struct net_device *dev);
static struct net_device_stats *get_stats(struct net_device *dev);
static int mii_ioctl(struct net_device *dev, struct ifreq *rq, int cmd);
static int  netdev_close(struct net_device *dev);



/* A list of our installed devices, for removing the driver module. */
static struct net_device *root_net_dev = NULL;

#ifndef MODULE
int starfire_probe(struct net_device *dev)
{
	if (pci_drv_register(&starfire_drv_id, dev) < 0)
		return -ENODEV;
	printk(KERN_INFO "%s" KERN_INFO "%s", version1, version2);
	return 0;
}
#endif

static void *starfire_probe1(struct pci_dev *pdev, void *init_dev,
							 long ioaddr, int irq, int chip_idx, int card_idx)
{
	struct net_device *dev;
	struct netdev_private *np;
	void *priv_mem;
	int i, option = card_idx < MAX_UNITS ? options[card_idx] : 0;

	dev = init_etherdev(init_dev, 0);
	if (!dev)
		return NULL;

	printk(KERN_INFO "%s: %s at 0x%lx, ",
		   dev->name, pci_id_tbl[chip_idx].name, ioaddr);

	/* Serial EEPROM reads are hidden by the hardware. */
	for (i = 0; i < 6; i++)
		dev->dev_addr[i] = readb(ioaddr + EEPROMCtrl + 20-i);
	for (i = 0; i < 5; i++)
		printk("%2.2x:", dev->dev_addr[i]);
	printk("%2.2x, IRQ %d.\n", dev->dev_addr[i], irq);

	/* Make certain elements e.g. descriptor lists are aligned. */
	priv_mem = kmalloc(sizeof(*np) + PRIV_ALIGN, GFP_KERNEL);
	/* Check for the very unlikely case of no memory. */
	if (priv_mem == NULL)
		return NULL;

	/* Reset the chip to erase previous misconfiguration. */
	writel(ChipResetCmd, ioaddr + PCIDeviceConfig);

	dev->base_addr = ioaddr;
	dev->irq = irq;

	dev->priv = np = (void *)(((long)priv_mem + PRIV_ALIGN) & ~PRIV_ALIGN);
	memset(np, 0, sizeof(*np));
	np->priv_addr = priv_mem;

	np->next_module = root_net_dev;
	root_net_dev = dev;

	np->pci_dev = pdev;
	np->chip_id = chip_idx;
	np->drv_flags = pci_id_tbl[chip_idx].drv_flags;
	np->msg_level = (1 << debug) - 1;
	np->rx_copybreak = rx_copybreak;
	np->max_interrupt_work = max_interrupt_work;
	np->multicast_filter_limit = multicast_filter_limit;

	if (dev->mem_start)
		option = dev->mem_start;

	if (card_idx < MAX_UNITS  &&  full_duplex[card_idx] > 0)
		np->full_duplex = 1;

	if (np->full_duplex) {
		if (np->msg_level & NETIF_MSG_PROBE)
			printk(KERN_INFO "%s: Set to forced full duplex, autonegotiation"
				   " disabled.\n", dev->name);
		np->medialock = 1;
	}

	/* The chip-specific entries in the device structure. */
	dev->open = &netdev_open;
	dev->hard_start_xmit = &start_tx;
	dev->stop = &netdev_close;
	dev->get_stats = &get_stats;
	dev->set_multicast_list = &set_rx_mode;
	dev->do_ioctl = &mii_ioctl;
	dev->change_mtu = &change_mtu;

	if (np->drv_flags & CanHaveMII) {
		int phy, phy_idx = 0;
		for (phy = 0; phy < 32 && phy_idx < 4; phy++) {
			int mii_status = mdio_read(dev, phy, 1);
			if (mii_status != 0xffff  &&  mii_status != 0x0000) {
				np->phys[phy_idx++] = phy;
				np->advertising = mdio_read(dev, phy, 4);
				if (np->msg_level & NETIF_MSG_PROBE)
					printk(KERN_INFO "%s: MII PHY found at address %d, status "
						   "0x%4.4x advertising %4.4x.\n",
						   dev->name, phy, mii_status, np->advertising);
			}
		}
		np->mii_cnt = phy_idx;
	}

	/* Force the media type after detecting the transceiver. */
	if (option > 0) {
		if (option & 0x220)
			np->full_duplex = 1;
		np->default_port = option & 0x3ff;
		if (np->default_port & 0x330) {
			np->medialock = 1;
			if (np->msg_level & NETIF_MSG_PROBE)
				printk(KERN_INFO "  Forcing %dMbs %s-duplex operation.\n",
					   (option & 0x300 ? 100 : 10),
					   (np->full_duplex ? "full" : "half"));
			mdio_write(dev, np->phys[0], 0,
					   ((option & 0x300) ? 0x2000 : 0) | 	/* 100mbps? */
					   (np->full_duplex ? 0x0100 : 0)); /* Full duplex? */
		}
	}

	return dev;
}


/* Read the MII Management Data I/O (MDIO) interfaces. */

static int mdio_read(struct net_device *dev, int phy_id, int location)
{
	long mdio_addr = dev->base_addr + MIICtrl + (phy_id<<7) + (location<<2);
	int result, boguscnt=1000;
	/* ??? Should we add a busy-wait here? */
	do
		result = readl(mdio_addr);
	while ((result & 0xC0000000) != 0x80000000 && --boguscnt >= 0);
	return result & 0xffff;
}

static void mdio_write(struct net_device *dev, int phy_id, int location, int value)
{
	long mdio_addr = dev->base_addr + MIICtrl + (phy_id<<7) + (location<<2);
	writel(value, mdio_addr);
	/* The busy-wait will occur before a read. */
	return;
}


static int netdev_open(struct net_device *dev)
{
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	long ioaddr = dev->base_addr;
	int i;

	MOD_INC_USE_COUNT;

	if (request_irq(dev->irq, &intr_handler, SA_SHIRQ, dev->name, dev)) {
		MOD_DEC_USE_COUNT;
		return -EAGAIN;
	}

	/* We have no reports that indicate we need to reset the chip.
	   But to be on the safe side... */
	/* Disable the Rx and Tx, and reset the chip. */
	writel(0, ioaddr + GenCtrl);
	writel(ChipResetCmd, ioaddr + PCIDeviceConfig);
	if (np->msg_level & NETIF_MSG_IFUP)
		printk(KERN_DEBUG "%s: netdev_open() irq %d.\n",
			   dev->name, dev->irq);
	/* Allocate the various queues, failing gracefully. */
	if (np->tx_done_q == 0)
		np->tx_done_q = (struct tx_done_report *)get_free_page(GFP_KERNEL);
	if (np->rx_done_q == 0)
		np->rx_done_q = (struct rx_done_desc *)get_free_page(GFP_KERNEL);
	if (np->tx_ring == 0)
		np->tx_ring = (struct starfire_tx_desc *)get_free_page(GFP_KERNEL);
	if (np->rx_ring == 0)
		np->rx_ring = (struct starfire_rx_desc *)get_free_page(GFP_KERNEL);
	if (np->tx_done_q == 0  ||  np->rx_done_q == 0
		|| np->rx_ring == 0 ||  np->tx_ring == 0) {
		/* Retain the pages to increase our chances next time. */
		MOD_DEC_USE_COUNT;
		return -ENOMEM;
	}

	init_ring(dev);
	/* Set the size of the Rx buffers. */
	writel((np->rx_buf_sz<<16) | 0xA000, ioaddr + RxDescQCtrl);

	/* Set Tx descriptor to type 1 and padding to 0 bytes. */
	writel(0x02000401, ioaddr + TxDescCtrl);

#if defined(STARFIRE_ADDR_64BITS)
	writel(virt_to_bus(np->rx_ring) >> 32, ioaddr + RxDescQHiAddr);
	writel(virt_to_bus(np->tx_ring) >> 32, ioaddr + TxRingHiAddr);
#else
	writel(0, ioaddr + RxDescQHiAddr);
	writel(0, ioaddr + TxRingHiAddr);
	writel(0, ioaddr + CompletionHiAddr);
#endif
	writel(virt_to_bus(np->rx_ring), ioaddr + RxDescQAddr);
	writel(virt_to_bus(np->tx_ring), ioaddr + TxRingPtr);

	writel(virt_to_bus(np->tx_done_q), ioaddr + TxCompletionAddr);
	writel(virt_to_bus(np->rx_done_q), ioaddr + RxCompletionAddr);

	if (np->msg_level & NETIF_MSG_IFUP)
		printk(KERN_DEBUG "%s:  Filling in the station address.\n", dev->name);

	/* Fill both the unused Tx SA register and the Rx perfect filter. */
	for (i = 0; i < 6; i++)
		writeb(dev->dev_addr[i], ioaddr + StationAddr + 5-i);
	for (i = 0; i < 16; i++) {
		u16 *eaddrs = (u16 *)dev->dev_addr;
		long setup_frm = ioaddr + 0x56000 + i*16;
		writew(cpu_to_be16(eaddrs[2]), setup_frm); setup_frm += 4;
		writew(cpu_to_be16(eaddrs[1]), setup_frm); setup_frm += 4;
		writew(cpu_to_be16(eaddrs[0]), setup_frm); setup_frm += 8;
	}

	/* Initialize other registers. */
	/* Configure the PCI bus bursts and FIFO thresholds. */
	np->tx_mode = 0;			/* Initialized when TxMode set. */
	np->tx_threshold = 4;
	writel(np->tx_threshold, ioaddr + TxThreshold);
	writel(interrupt_mitigation, ioaddr + IntrTimerCtrl);

	if (dev->if_port == 0)
		dev->if_port = np->default_port;

	if (np->msg_level & NETIF_MSG_IFUP)
		printk(KERN_DEBUG "%s:  Setting the Rx and Tx modes.\n", dev->name);
	set_rx_mode(dev);

	np->advertising = mdio_read(dev, np->phys[0], 4);
	check_duplex(dev);
	netif_start_tx_queue(dev);

	/* Set the interrupt mask and enable PCI interrupts. */
	np->intr_enable = IntrRxDone | IntrRxEmpty | IntrRxPCIErr |
		IntrTxDone | IntrTxEmpty | IntrTxPCIErr |
		StatsMax | LinkChange | IntrNormalSummary | IntrAbnormalSummary
		| 0x0010;
	writel(np->intr_enable, ioaddr + IntrEnable);
	writel(PCIIntEnb | readl(ioaddr + PCIDeviceConfig),
		   ioaddr + PCIDeviceConfig);

	/* Enable the Rx and Tx units. */
	writel(TxEnable|RxEnable, ioaddr + GenCtrl);

	if (np->msg_level & NETIF_MSG_IFUP)
		printk(KERN_DEBUG "%s: Done netdev_open().\n",
			   dev->name);

	/* Set the timer to check for link beat. */
	init_timer(&np->timer);
	np->timer.expires = jiffies + 3*HZ;
	np->timer.data = (unsigned long)dev;
	np->timer.function = &netdev_timer;				/* timer handler */
	add_timer(&np->timer);

	return 0;
}

/* The starfire can handle frame sizes up to 64KB, but we arbitrarily
 * limit the size.
 */
static int change_mtu(struct net_device *dev, int new_mtu)
{
	if ((new_mtu < 68) || (new_mtu > 17268))
		return -EINVAL;
	if (netif_running(dev))
		return -EBUSY;
	dev->mtu = new_mtu;
	return 0;
}

static void check_duplex(struct net_device *dev)
{
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	long ioaddr = dev->base_addr;
	int new_tx_mode;

	new_tx_mode = 0x0C04 | (np->tx_flowctrl ? 0x0800:0)
		| (np->rx_flowctrl ? 0x0400:0);
	if (np->medialock) {
		if (np->full_duplex)
			new_tx_mode |= 2;
	} else {
		int mii_reg5 = mdio_read(dev, np->phys[0], 5);
		int negotiated = mii_reg5 & np->advertising;
		int duplex = (negotiated & 0x0100) || (negotiated & 0x01C0) == 0x0040;
		if (duplex)
			new_tx_mode |= 2;
		if (np->full_duplex != duplex) {
			np->full_duplex = duplex;
			if (np->msg_level & NETIF_MSG_LINK)
				printk(KERN_INFO "%s: Setting %s-duplex based on MII #%d"
					   " negotiated capability %4.4x.\n", dev->name,
					   duplex ? "full" : "half", np->phys[0], negotiated);
		}
	}
	if (new_tx_mode != np->tx_mode) {
		np->tx_mode = new_tx_mode;
		writel(np->tx_mode | 0x8000, ioaddr + TxMode);
		writel(np->tx_mode, ioaddr + TxMode);
	}
}

/* Check for duplex changes, but mostly check for failures. */
static void netdev_timer(unsigned long data)
{
	struct net_device *dev = (struct net_device *)data;
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	long ioaddr = dev->base_addr;
	int status = readl(ioaddr + IntrStatus);
	static long last_msg = 0;

	/* Normally we check only every few seconds. */
	np->timer.expires = jiffies + 60*HZ;

	if (np->msg_level & NETIF_MSG_TIMER) {
		printk(KERN_DEBUG "%s: Media selection timer tick, status %8.8x.\n",
			   dev->name, status);
	}

	/* Check for a missing chip or failed interrupt line.
	 * The latter may be falsely triggered, so we check twice. */
	if (status == 0xffffffff) {
		if (jiffies - last_msg > 10*HZ) {
			last_msg = jiffies;
			printk(KERN_ERR "%s: The Starfire chip is missing!\n",
				   dev->name);
		}
	} else if (np->polling) {
		if (status & IntrPCIPin) {
			intr_handler(dev->irq, dev, 0);
			if (jiffies - last_msg > 10*HZ) {
				printk(KERN_ERR "%s: IRQ %d is still blocked!\n",
					   dev->name, dev->irq);
				last_msg = jiffies;
			}
		} else if (jiffies - last_msg > 10*HZ)
			np->polling = 0;
		np->timer.expires = jiffies + 2;
	} else if (status & IntrPCIPin) {
		int new_status = readl(ioaddr + IntrStatus);
		/* Bogus hardware IRQ mapping: Fake an interrupt handler call. */
		if (new_status & IntrPCIPin) {
			printk(KERN_ERR "%s: IRQ %d is not raising an interrupt! "
				   "Status %8.8x/%8.8x.  \n",
				   dev->name, dev->irq, status, new_status);
			intr_handler(dev->irq, dev, 0);
			np->timer.expires = jiffies + 2;
			np->polling = 1;
		}
	} else if (netif_queue_paused(dev)  &&
			   np->cur_tx - np->dirty_tx > 1  &&
			   (jiffies - dev->trans_start) > TX_TIMEOUT) {
		/* This will not catch tbusy incorrectly set when the queue is empty,
		 * but that state should never occur. */
		tx_timeout(dev);
	}

	check_duplex(dev);

	add_timer(&np->timer);
}

static void tx_timeout(struct net_device *dev)
{
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	long ioaddr = dev->base_addr;

	printk(KERN_WARNING "%s: Transmit timed out, status %8.8x,"
		   " resetting...\n", dev->name, (int)readl(ioaddr + IntrStatus));

#if defined(__i386__)
	if (np->msg_level & NETIF_MSG_TX_ERR) {
		int i;
		printk("\n" KERN_DEBUG "  Tx ring %p: ", np->tx_ring);
		for (i = 0; i < TX_RING_SIZE; i++)
			printk(" %4.4x", np->tx_ring[i].status);
		printk("\n" KERN_DEBUG "  Rx ring %p: ", np->rx_ring);
		for (i = 0; i < RX_RING_SIZE; i++)
			printk(" %8.8x", (unsigned int)np->rx_ring[i].rxaddr);
		printk("\n");
	}
#endif

	/* If a specific problem is reported, reinitialize the hardware here. */
	dev->if_port = 0;
	/* Stop and restart the chip's Tx processes . */
	writel(0, ioaddr + GenCtrl);
	/* Enable the Rx and Tx units. */
	writel(TxEnable|RxEnable, ioaddr + GenCtrl);

	dev->trans_start = jiffies;
	np->stats.tx_errors++;
	return;
}


/* Initialize the Rx and Tx rings, along with various 'dev' bits. */
static void init_ring(struct net_device *dev)
{
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	int i;

	np->tx_full = 0;
	np->cur_rx = np->cur_tx = 0;
	np->dirty_rx = np->rx_done = np->dirty_tx = np->tx_done = 0;

	np->rx_buf_sz = (dev->mtu <= 1522 ? PKT_BUF_SZ :
					 (dev->mtu + 14 + 3) & ~3);	/* Round to word. */

	/* Fill in the Rx buffers.  Handle allocation failure gracefully. */
	for (i = 0; i < RX_RING_SIZE; i++) {
		struct sk_buff *skb = dev_alloc_skb(np->rx_buf_sz);
		np->rx_skbuff[i] = skb;
		if (skb == NULL)
			break;
		skb->dev = dev;			/* Mark as being used by this device. */
		/* Grrr, we cannot offset to correctly align the IP header. */
		np->rx_ring[i].rxaddr =
			virt_to_le32desc(skb->tail) | cpu_to_le32(RxDescValid);
	}
	writew(i - 1, dev->base_addr + RxDescQIdx);
	np->dirty_rx = (unsigned int)(i - RX_RING_SIZE);

	/* Clear the remainder of the Rx buffer ring. */
	for (  ; i < RX_RING_SIZE; i++) {
		np->rx_ring[i].rxaddr = 0;
		np->rx_skbuff[i] = 0;
	}
	/* Mark the last entry as wrapping the ring. */
	np->rx_ring[i-1].rxaddr |= cpu_to_le32(RxDescEndRing);

	/* Clear the completion rings. */
	for (i = 0; i < DONE_Q_SIZE; i++) {
		np->rx_done_q[i].status = 0;
		np->tx_done_q[i].status = 0;
	}

	for (i = 0; i < TX_RING_SIZE; i++) {
		np->tx_skbuff[i] = 0;
		np->tx_ring[i].status = 0;
	}
	return;
}

static int start_tx(struct sk_buff *skb, struct net_device *dev)
{
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	unsigned entry;

	/* Block a timer-based transmit from overlapping.  This happens when
	   packets are presumed lost, and we use this check the Tx status. */
	if (netif_pause_tx_queue(dev) != 0) {
		/* This watchdog code is redundant with the media monitor timer. */
		if (jiffies - dev->trans_start > TX_TIMEOUT)
			tx_timeout(dev);
		return 1;
	}

	/* Caution: the write order is important here, set the field
	   with the "ownership" bits last. */

	/* Calculate the next Tx descriptor entry. */
	entry = np->cur_tx % TX_RING_SIZE;

	np->tx_skbuff[entry] = skb;

	np->tx_ring[entry].addr = virt_to_le32desc(skb->data);
	/* Add  "| TxDescIntr" to generate Tx-done interrupts. */
	np->tx_ring[entry].status = cpu_to_le32(skb->len | TxDescID);
#if 1
	if (entry >= TX_RING_SIZE-1) {		 /* Wrap ring */
		np->tx_ring[entry].status |= cpu_to_le32(TxRingWrap | TxDescIntr);
		entry = -1;
	}
#endif

	/* On some architectures better performance results by explicitly
	   flushing cache lines: pci_flush_virt(skb->data, skb->len); */

	np->cur_tx++;
	/* Update the producer index. */
	writel(++entry, dev->base_addr + TxProducerIdx);

	/* cf. using TX_QUEUE_LEN instead of TX_RING_SIZE here. */
	if (np->cur_tx - np->dirty_tx >= TX_RING_SIZE - 1) {
		np->tx_full = 1;
		/* Check for the rare case of a just-cleared queue. */
		if (np->cur_tx - (volatile unsigned int)np->dirty_tx
			< TX_RING_SIZE - 2) {
			np->tx_full = 0;
			netif_unpause_tx_queue(dev);
		} else
			netif_stop_tx_queue(dev);
	} else
		netif_unpause_tx_queue(dev);		/* Typical path */

	dev->trans_start = jiffies;

	if (np->msg_level & NETIF_MSG_TX_QUEUED) {
		printk(KERN_DEBUG "%s: Tx frame #%d slot %d  %8.8x %8.8x.\n",
			   dev->name, np->cur_tx, entry,
			   np->tx_ring[entry].status, np->tx_ring[entry].addr);
	}
	return 0;
}

/* The interrupt handler does all of the Rx thread work and cleans up
   after the Tx thread. */
static void intr_handler(int irq, void *dev_instance, struct pt_regs *rgs)
{
	struct net_device *dev = (struct net_device *)dev_instance;
	struct netdev_private *np;
	long ioaddr;
	int boguscnt;

#ifndef final_version			/* Can never occur. */
	if (dev == NULL) {
		printk (KERN_ERR "Netdev interrupt handler(): IRQ %d for unknown "
				"device.\n", irq);
		return;
	}
#endif

	ioaddr = dev->base_addr;
	np = (struct netdev_private *)dev->priv;
	boguscnt = np->max_interrupt_work;

	do {
		u32 intr_status = readl(ioaddr + IntrClear);

		if (np->msg_level & NETIF_MSG_INTR)
			printk(KERN_DEBUG "%s: Interrupt status %4.4x.\n",
				   dev->name, intr_status);

		if (intr_status == 0 || intr_status == 0xffffffff)
			break;

		if (intr_status & IntrRxDone)
			netdev_rx(dev);

		/* Scavenge the skbuff list based on the Tx-done queue.
		   There are redundant checks here that may be cleaned up
		   after the driver has proven to be reliable. */
		{
			int consumer = readl(ioaddr + TxConsumerIdx);
			int tx_status;
			if (np->msg_level & NETIF_MSG_INTR)
				printk(KERN_DEBUG "%s: Tx Consumer index is %d.\n",
					   dev->name, consumer);
#if 0
			if (np->tx_done >= 250  || np->tx_done == 0)
				printk(KERN_DEBUG "%s: Tx completion entry %d is %8.8x, "
					   "%d is %8.8x.\n", dev->name,
					   np->tx_done, np->tx_done_q[np->tx_done].status,
					   (np->tx_done+1) & (DONE_Q_SIZE-1),
					   np->tx_done_q[(np->tx_done+1)&(DONE_Q_SIZE-1)].status);
#endif
			while ((tx_status = cpu_to_le32(np->tx_done_q[np->tx_done].status))
				   != 0) {
				if (np->msg_level & NETIF_MSG_TX_DONE)
					printk(KERN_DEBUG "%s: Tx completion entry %d is %8.8x.\n",
						   dev->name, np->tx_done, tx_status);
				if ((tx_status & 0xe0000000) == 0xa0000000) {
					np->stats.tx_packets++;
				} else if ((tx_status & 0xe0000000) == 0x80000000) {
					u16 entry = tx_status; 		/* Implicit truncate */
					entry >>= 3;
					/* Scavenge the descriptor. */
					if (np->tx_skbuff[entry]) {
						dev_free_skb_irq(np->tx_skbuff[entry]);
					} else
						printk(KERN_WARNING "%s: Null skbuff at entry %d!!!\n",
							   dev->name, entry);
					np->tx_skbuff[entry] = 0;
					np->dirty_tx++;
				}
				np->tx_done_q[np->tx_done].status = 0;
				np->tx_done = (np->tx_done+1) & (DONE_Q_SIZE-1);
			}
			writew(np->tx_done, ioaddr + CompletionQConsumerIdx + 2);
		}
		if (np->tx_full && np->cur_tx - np->dirty_tx < TX_RING_SIZE - 4) {
			/* The ring is no longer full, allow new TX entries. */
			np->tx_full = 0;
			netif_resume_tx_queue(dev);
		}

		/* Abnormal error summary/uncommon events handlers. */
		if (intr_status & IntrAbnormalSummary)
			netdev_error(dev, intr_status);

		if (--boguscnt < 0) {
			printk(KERN_WARNING "%s: Too much work at interrupt, "
				   "status=0x%4.4x.\n",
				   dev->name, intr_status);
			writel(0x0021, ioaddr + IntrTimerCtrl);
			break;
		}
	} while (1);

	if (np->msg_level & NETIF_MSG_INTR)
		printk(KERN_DEBUG "%s: exiting interrupt, status=%#4.4x.\n",
			   dev->name, (int)readl(ioaddr + IntrStatus));

	return;
}

/* This routine is logically part of the interrupt handler, but separated
   for clarity and better register allocation. */
static int netdev_rx(struct net_device *dev)
{
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	int boguscnt = np->dirty_rx + RX_RING_SIZE - np->cur_rx;
	u32 desc_status;

	if (np->rx_done_q == 0) {
		printk(KERN_ERR "%s:  rx_done_q is NULL!  rx_done is %d. %p.\n",
			   dev->name, np->rx_done, np->tx_done_q);
		return 0;
	}

	/* If EOP is set on the next entry, it's a new packet. Send it up. */
	while ((desc_status = le32_to_cpu(np->rx_done_q[np->rx_done].status)) != 0) {
		if (np->msg_level & NETIF_MSG_RX_STATUS)
			printk(KERN_DEBUG "  netdev_rx() status of %d was %8.8x.\n",
				   np->rx_done, desc_status);
		if (--boguscnt < 0)
			break;
		if ( ! (desc_status & RxOK)) {
			/* There was a error. */
			if (np->msg_level & NETIF_MSG_RX_ERR)
				printk(KERN_DEBUG "  netdev_rx() Rx error was %8.8x.\n",
					   desc_status);
			np->stats.rx_errors++;
			if (desc_status & RxFIFOErr)
				np->stats.rx_fifo_errors++;
		} else {
			struct sk_buff *skb;
			u16 pkt_len = desc_status;			/* Implicitly Truncate */
			int entry = (desc_status >> 16) & 0x7ff;

#ifndef final_version
			if (np->msg_level & NETIF_MSG_RX_STATUS)
				printk(KERN_DEBUG "  netdev_rx() normal Rx pkt length %d"
					   ", bogus_cnt %d.\n",
					   pkt_len, boguscnt);
#endif
			/* Check if the packet is long enough to accept without copying
			   to a minimally-sized skbuff. */
			if (pkt_len < rx_copybreak
				&& (skb = dev_alloc_skb(pkt_len + 2)) != NULL) {
				skb->dev = dev;
				skb_reserve(skb, 2);	/* 16 byte align the IP header */
#if HAS_IP_COPYSUM			/* Call copy + cksum if available. */
				eth_copy_and_sum(skb, np->rx_skbuff[entry]->tail, pkt_len, 0);
				skb_put(skb, pkt_len);
#else
				memcpy(skb_put(skb, pkt_len), np->rx_skbuff[entry]->tail,
					   pkt_len);
#endif
			} else {
				char *temp = skb_put(skb = np->rx_skbuff[entry], pkt_len);
				np->rx_skbuff[entry] = NULL;
#ifndef final_version				/* Remove after testing. */
				if (le32desc_to_virt(np->rx_ring[entry].rxaddr & ~3) != temp)
					printk(KERN_ERR "%s: Internal fault: The skbuff addresses "
						   "do not match in netdev_rx: %p vs. %p / %p.\n",
						   dev->name,
						   le32desc_to_virt(np->rx_ring[entry].rxaddr),
						   skb->head, temp);
#endif
			}
			skb->protocol = eth_type_trans(skb, dev);
#ifdef full_rx_status
			if (np->rx_done_q[np->rx_done].status2 & cpu_to_le32(0x01000000))
				skb->ip_summed = CHECKSUM_UNNECESSARY;
#endif
			netif_rx(skb);
			dev->last_rx = jiffies;
			np->stats.rx_packets++;
		}
		np->cur_rx++;
		np->rx_done_q[np->rx_done].status = 0;
		np->rx_done = (np->rx_done + 1) & (DONE_Q_SIZE-1);
	}
	writew(np->rx_done, dev->base_addr + CompletionQConsumerIdx);

	/* Refill the Rx ring buffers. */
	for (; np->cur_rx - np->dirty_rx > 0; np->dirty_rx++) {
		struct sk_buff *skb;
		int entry = np->dirty_rx % RX_RING_SIZE;
		if (np->rx_skbuff[entry] == NULL) {
			skb = dev_alloc_skb(np->rx_buf_sz);
			np->rx_skbuff[entry] = skb;
			if (skb == NULL)
				break;				/* Better luck next round. */
			skb->dev = dev;			/* Mark as being used by this device. */
			np->rx_ring[entry].rxaddr =
				virt_to_le32desc(skb->tail) | cpu_to_le32(RxDescValid);
		}
		if (entry == RX_RING_SIZE - 1)
			np->rx_ring[entry].rxaddr |= cpu_to_le32(RxDescEndRing);
		/* We could defer this until later... */
		writew(entry, dev->base_addr + RxDescQIdx);
	}

	if ((np->msg_level & NETIF_MSG_RX_STATUS)
		|| memcmp(np->pad0, np->pad0 + 1, sizeof(np->pad0) -1))
		printk(KERN_DEBUG "  exiting netdev_rx() status of %d was %8.8x %d.\n",
			   np->rx_done, desc_status,
			   memcmp(np->pad0, np->pad0 + 1, sizeof(np->pad0) -1));

	return 0;
}

static void netdev_error(struct net_device *dev, int intr_status)
{
	struct netdev_private *np = (struct netdev_private *)dev->priv;

	if (intr_status & LinkChange) {
		int phy_num = np->phys[0];
		if (np->msg_level & NETIF_MSG_LINK)
			printk(KERN_NOTICE "%s: Link changed: Autonegotiation advertising"
				   " %4.4x  partner %4.4x.\n", dev->name,
				   mdio_read(dev, phy_num, 4),
				   mdio_read(dev, phy_num, 5));
		/* Clear sticky bit. */
		mdio_read(dev, phy_num, 1);
		/* If link beat has returned... */
		if (mdio_read(dev, phy_num, 1) & 0x0004)
			netif_link_up(dev);
		else
			netif_link_down(dev);
		check_duplex(dev);
	}
	if (intr_status & StatsMax) {
		get_stats(dev);
	}
	/* Came close to underrunning the Tx FIFO, increase threshold. */
	if (intr_status & IntrTxDataLow)
		writel(++np->tx_threshold, dev->base_addr + TxThreshold);
	/* Ingore expected normal events, and handled abnormal events. */
	if ((intr_status &
		 ~(IntrAbnormalSummary|LinkChange|StatsMax|IntrTxDataLow| 0xFF01))
		&& (np->msg_level & NETIF_MSG_DRV))
		printk(KERN_ERR "%s: Something Wicked happened! %4.4x.\n",
			   dev->name, intr_status);
	/* Hmmmmm, it's not clear how to recover from PCI faults. */
	if (intr_status & IntrTxPCIErr)
		np->stats.tx_fifo_errors++;
	if (intr_status & IntrRxPCIErr)
		np->stats.rx_fifo_errors++;
}

static struct net_device_stats *get_stats(struct net_device *dev)
{
	long ioaddr = dev->base_addr;
	struct netdev_private *np = (struct netdev_private *)dev->priv;

	/* This adapter architecture needs no SMP locks. */
#if LINUX_VERSION_CODE > 0x20119
	np->stats.tx_bytes = readl(ioaddr + 0x57010);
	np->stats.rx_bytes = readl(ioaddr + 0x57044);
#endif
	np->stats.tx_packets = readl(ioaddr + 0x57000);
	np->stats.tx_aborted_errors =
		readl(ioaddr + 0x57024) + readl(ioaddr + 0x57028);
	np->stats.tx_window_errors = readl(ioaddr + 0x57018);
	np->stats.collisions = readl(ioaddr + 0x57004) + readl(ioaddr + 0x57008);

	/* The chip only need report frame silently dropped. */
	np->stats.rx_dropped	   += readw(ioaddr + RxDMAStatus);
	writew(0, ioaddr + RxDMAStatus);
	np->stats.rx_crc_errors	   = readl(ioaddr + 0x5703C);
	np->stats.rx_frame_errors = readl(ioaddr + 0x57040);
	np->stats.rx_length_errors = readl(ioaddr + 0x57058);
	np->stats.rx_missed_errors = readl(ioaddr + 0x5707C);

	return &np->stats;
}

/* The little-endian AUTODIN II ethernet CRC calculations.
   A big-endian version is also available.
   This is slow but compact code.  Do not use this routine for bulk data,
   use a table-based routine instead.
   This is common code and should be moved to net/core/crc.c.
   Chips may use the upper or lower CRC bits, and may reverse and/or invert
   them.  Select the endian-ness that results in minimal calculations.
*/
static unsigned const ethernet_polynomial_le = 0xedb88320U;
static inline unsigned ether_crc_le(int length, unsigned char *data)
{
	unsigned int crc = ~0;	/* Initial value. */
	while(--length >= 0) {
		unsigned char current_octet = *data++;
		int bit;
		for (bit = 8; --bit >= 0; current_octet >>= 1) {
			if ((crc ^ current_octet) & 1) {
				crc >>= 1;
				crc ^= ethernet_polynomial_le;
			} else
				crc >>= 1;
		}
	}
	return crc;
}

static void set_rx_mode(struct net_device *dev)
{
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	long ioaddr = dev->base_addr;
	u32 rx_mode;
	struct dev_mc_list *mclist;
	int i;

	if (dev->flags & IFF_PROMISC) {			/* Set promiscuous. */
		/* Unconditionally log net taps. */
		printk(KERN_NOTICE "%s: Promiscuous mode enabled.\n", dev->name);
		rx_mode = AcceptBroadcast|AcceptAllMulticast|AcceptAll|AcceptMyPhys;
	} else if ((dev->mc_count > np->multicast_filter_limit)
			   ||  (dev->flags & IFF_ALLMULTI)) {
		/* Too many to match, or accept all multicasts. */
		rx_mode = AcceptBroadcast|AcceptAllMulticast|AcceptMyPhys;
	} else if (dev->mc_count <= 15) {
		/* Use the 16 element perfect filter. */
		long filter_addr = ioaddr + 0x56000 + 1*16;
		for (i = 1, mclist = dev->mc_list; mclist  &&  i <= dev->mc_count;
			 i++, mclist = mclist->next) {
			u16 *eaddrs = (u16 *)mclist->dmi_addr;
			writew(cpu_to_be16(eaddrs[2]), filter_addr); filter_addr += 4;
			writew(cpu_to_be16(eaddrs[1]), filter_addr); filter_addr += 4;
			writew(cpu_to_be16(eaddrs[0]), filter_addr); filter_addr += 8;
		}
		while (i++ < 16) {
			writew(0xffff, filter_addr); filter_addr += 4;
			writew(0xffff, filter_addr); filter_addr += 4;
			writew(0xffff, filter_addr); filter_addr += 8;
		}
		rx_mode = AcceptBroadcast | AcceptMyPhys;
	} else {
		/* Must use a multicast hash table. */
		long filter_addr;
		u16 mc_filter[32];			/* Multicast hash filter */

		memset(mc_filter, 0, sizeof(mc_filter));
		for (i = 0, mclist = dev->mc_list; mclist && i < dev->mc_count;
			 i++, mclist = mclist->next) {
			set_bit(ether_crc_le(ETH_ALEN, mclist->dmi_addr) >> 23, mc_filter);
		}
		/* Clear the perfect filter list. */
		filter_addr = ioaddr + 0x56000 + 1*16;
		for (i = 1; i < 16; i++) {
			writew(0xffff, filter_addr); filter_addr += 4;
			writew(0xffff, filter_addr); filter_addr += 4;
			writew(0xffff, filter_addr); filter_addr += 8;
		}
		for (filter_addr=ioaddr + 0x56100, i=0; i < 32; filter_addr+= 16, i++){
			np->mc_filter[i] = mc_filter[i];
			writew(mc_filter[i], filter_addr);
		}
		rx_mode = AcceptBroadcast | AcceptMulticast | AcceptMyPhys;
	}
	writel(rx_mode, ioaddr + RxFilterMode);
}

/*
  Handle user-level ioctl() calls.
  We must use two numeric constants as the key because some clueless person
  changed the value for the symbolic name.
*/
static int mii_ioctl(struct net_device *dev, struct ifreq *rq, int cmd)
{
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	u16 *data = (u16 *)&rq->ifr_data;
	u32 *data32 = (void *)&rq->ifr_data;

	switch(cmd) {
	case 0x8947: case 0x89F0:
		/* SIOCGMIIPHY: Get the address of the PHY in use. */
		data[0] = np->phys[0] & 0x1f;
		/* Fall Through */
	case 0x8948: case 0x89F1:
		/* SIOCGMIIREG: Read the specified MII register. */
		data[3] = mdio_read(dev, data[0] & 0x1f, data[1] & 0x1f);
		return 0;
	case 0x8949: case 0x89F2:
		/* SIOCSMIIREG: Write the specified MII register */
		if (!capable(CAP_NET_ADMIN))
			return -EPERM;
		if (data[0] == np->phys[0]) {
			u16 value = data[2];
			switch (data[1]) {
			case 0:
				/* Check for autonegotiation on or reset. */
				np->medialock = (value & 0x9000) ? 0 : 1;
				if (np->medialock)
					np->full_duplex = (value & 0x0100) ? 1 : 0;
				break;
			case 4: np->advertising = value; break;
			}
			check_duplex(dev);
		}
		mdio_write(dev, data[0] & 0x1f, data[1] & 0x1f, data[2]);
		return 0;
	case SIOCGPARAMS:
		data32[0] = np->msg_level;
		data32[1] = np->multicast_filter_limit;
		data32[2] = np->max_interrupt_work;
		data32[3] = np->rx_copybreak;
		return 0;
	case SIOCSPARAMS:
		if (!capable(CAP_NET_ADMIN))
			return -EPERM;
		np->msg_level = data32[0];
		np->multicast_filter_limit = data32[1];
		np->max_interrupt_work = data32[2];
		np->rx_copybreak = data32[3];
		return 0;
	default:
		return -EOPNOTSUPP;
	}
}

static int netdev_close(struct net_device *dev)
{
	long ioaddr = dev->base_addr;
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	int i;

	netif_stop_tx_queue(dev);

	if (np->msg_level & NETIF_MSG_IFDOWN) {
		printk(KERN_DEBUG "%s: Shutting down ethercard, Intr status %4.4x.\n",
			   dev->name, (int)readl(ioaddr + IntrStatus));
		printk(KERN_DEBUG "%s: Queue pointers were Tx %d / %d,  Rx %d / %d.\n",
			   dev->name, np->cur_tx, np->dirty_tx, np->cur_rx, np->dirty_rx);
	}

	/* Disable interrupts by clearing the interrupt mask. */
	writel(0, ioaddr + IntrEnable);

	/* Stop the chip's Tx and Rx processes. */
	writel(0, ioaddr + GenCtrl);

	del_timer(&np->timer);

#ifdef __i386__
	if (np->msg_level & NETIF_MSG_IFDOWN) {
		printk("\n"KERN_DEBUG"  Tx ring at %8.8x:\n",
			   (int)virt_to_bus(np->tx_ring));
		for (i = 0; i < 8 /* TX_RING_SIZE is huge! */; i++)
			printk(KERN_DEBUG " #%d desc. %8.8x %8.8x -> %8.8x.\n",
				   i, np->tx_ring[i].status, np->tx_ring[i].addr,
				   np->tx_done_q[i].status);
		printk(KERN_DEBUG "  Rx ring at %8.8x -> %p:\n",
			   (int)virt_to_bus(np->rx_ring), np->rx_done_q);
		if (np->rx_done_q)
			for (i = 0; i < 8 /* RX_RING_SIZE */; i++) {
				printk(KERN_DEBUG " #%d desc. %8.8x -> %8.8x\n",
					   i, np->rx_ring[i].rxaddr, np->rx_done_q[i].status);
		}
	}
#endif /* __i386__ debugging only */

	free_irq(dev->irq, dev);

	/* Free all the skbuffs in the Rx queue. */
	for (i = 0; i < RX_RING_SIZE; i++) {
		np->rx_ring[i].rxaddr = 0xBADF00D0; /* An invalid address. */
		if (np->rx_skbuff[i]) {
#if LINUX_VERSION_CODE < 0x20100
			np->rx_skbuff[i]->free = 1;
#endif
			dev_free_skb(np->rx_skbuff[i]);
		}
		np->rx_skbuff[i] = 0;
	}
	for (i = 0; i < TX_RING_SIZE; i++) {
		if (np->tx_skbuff[i])
			dev_free_skb(np->tx_skbuff[i]);
		np->tx_skbuff[i] = 0;
	}

	MOD_DEC_USE_COUNT;

	return 0;
}


static int starfire_pwr_event(void *dev_instance, int event)
{
	struct net_device *dev = dev_instance;
	struct netdev_private *np = (struct netdev_private *)dev->priv;
	long ioaddr = dev->base_addr;

	if (np->msg_level & NETIF_MSG_LINK)
		printk(KERN_DEBUG "%s: Handling power event %d.\n", dev->name, event);
	switch(event) {
	case DRV_ATTACH:
		MOD_INC_USE_COUNT;
		break;
	case DRV_SUSPEND:
		/* Disable interrupts, stop Tx and Rx. */
		writel(0x0000, ioaddr + IntrEnable);
		writel(0, ioaddr + GenCtrl);
		break;
	case DRV_RESUME:
		/* This is incomplete: we must factor start_chip() out of open(). */
		writel(np->tx_threshold, ioaddr + TxThreshold);
		writel(interrupt_mitigation, ioaddr + IntrTimerCtrl);
		set_rx_mode(dev);
		writel(np->intr_enable, ioaddr + IntrEnable);
		writel(TxEnable|RxEnable, ioaddr + GenCtrl);
		break;
	case DRV_DETACH: {
		struct net_device **devp, **next;
		if (dev->flags & IFF_UP) {
			/* Some, but not all, kernel versions close automatically. */
			dev_close(dev);
			dev->flags &= ~(IFF_UP|IFF_RUNNING);
		}
		unregister_netdev(dev);
		release_region(dev->base_addr, pci_id_tbl[np->chip_id].io_size);
#ifndef USE_IO_OPS
		iounmap((char *)dev->base_addr);
#endif
		for (devp = &root_net_dev; *devp; devp = next) {
			next = &((struct netdev_private *)(*devp)->priv)->next_module;
			if (*devp == dev) {
				*devp = *next;
				break;
			}
		}
		if (np->priv_addr)
			kfree(np->priv_addr);
		kfree(dev);
		MOD_DEC_USE_COUNT;
		break;
	}
	}

	return 0;
}


#ifdef MODULE
int init_module(void)
{
	if (debug >= NETIF_MSG_DRV)	/* Emit version even if no cards detected. */
		printk(KERN_INFO "%s" KERN_INFO "%s", version1, version2);
	if (pci_drv_register(&starfire_drv_id, NULL)) {
		printk(KERN_INFO " No Starfire adapters detected, driver not loaded.\n");
		return -ENODEV;
	}
	return 0;
}

void cleanup_module(void)
{
	struct net_device *next_dev;

	pci_drv_unregister(&starfire_drv_id);

	/* No need to check MOD_IN_USE, as sys_delete_module() checks. */
	while (root_net_dev) {
		struct netdev_private *np = (void *)(root_net_dev->priv);
		unregister_netdev(root_net_dev);
		iounmap((char *)(root_net_dev->base_addr));
		next_dev = np->next_module;
		if (np->tx_done_q) free_page((long)np->tx_done_q);
		if (np->rx_done_q) free_page((long)np->rx_done_q);
		if (np->priv_addr) kfree(np->priv_addr);
		kfree(root_net_dev);
		root_net_dev = next_dev;
	}
}

#endif  /* MODULE */

/*
 * Local variables:
 *  compile-command: "make KERNVER=`uname -r` starfire.o"
 *  compile-cmd: "gcc -DMODULE -Wall -Wstrict-prototypes -O6 -c starfire.c"
 *  simple-compile-command: "gcc -DMODULE -O6 -c starfire.c"
 *  c-indent-level: 4
 *  c-basic-offset: 4
 *  tab-width: 4
 * End:
 */