patch-2.4.7 linux/arch/cris/drivers/lpslave/e100lpslavenet.c

Next file: linux/arch/cris/drivers/parport.c
Previous file: linux/arch/cris/drivers/lpslave/e100lpslaveld
Back to the patch index
Back to the overall index

diff -u --recursive --new-file v2.4.6/linux/arch/cris/drivers/lpslave/e100lpslavenet.c linux/arch/cris/drivers/lpslave/e100lpslavenet.c
@@ -0,0 +1,1031 @@
+/* $Id: e100lpslavenet.c,v 1.2 2001/06/11 15:39:52 olof Exp $
+ *
+ * e100lpslavenet.c: A network driver for the ETRAX 100LX slave controller.
+ *
+ * Copyright (c) 1998-2001 Axis Communications AB.
+ *
+ * The outline of this driver comes from skeleton.c.
+ *
+ * $Log: e100lpslavenet.c,v $
+ * Revision 1.2  2001/06/11 15:39:52  olof
+ * Clean up and sync with ethernet.c rev 1.16. Increased reset time of slave.
+ *
+ * Revision 1.1  2001/06/06 08:56:26  olof
+ * Added support for slave Etrax defined by CONFIG_ETRAX_ETHERNET_LPSLAVE
+ *
+ */
+
+#include <linux/config.h>
+
+#include <linux/module.h>
+
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/delay.h>
+#include <linux/types.h>
+#include <linux/fcntl.h>
+#include <linux/interrupt.h>
+#include <linux/ptrace.h>
+#include <linux/ioport.h>
+#include <linux/in.h>
+#include <linux/slab.h>
+#include <linux/string.h>
+#include <linux/spinlock.h>
+#include <linux/errno.h>
+#include <linux/init.h>
+
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/skbuff.h>
+
+#include <asm/svinto.h>     /* DMA and register descriptions */
+#include <asm/io.h>         /* LED_* I/O functions */
+#include <asm/irq.h>
+#include <asm/dma.h>
+#include <asm/system.h>
+#include <asm/bitops.h>
+
+#include "e100lpslave.h"
+
+/* #define ETHDEBUG */
+#define D(x)
+
+/*
+ * The name of the card. Is used for messages and in the requests for
+ * io regions, irqs and dma channels
+ */
+
+static const char* cardname = "Etrax 100LX ethernet slave controller";
+
+/* A default ethernet address. Highlevel SW will set the real one later */
+
+static struct sockaddr default_mac = {
+	0,
+        { 0x00, 0x40, 0x8C, 0xCD, 0x00, 0x00 }
+};
+
+/* Information that need to be kept for each board. */
+struct net_local {
+	struct net_device_stats stats;
+
+	/* Tx control lock.  This protects the transmit buffer ring
+	 * state along with the "tx full" state of the driver.  This
+	 * means all netif_queue flow control actions are protected
+	 * by this lock as well.
+	 */
+	spinlock_t lock;
+};
+
+/* Dma descriptors etc. */
+
+#define RX_BUF_SIZE 32768
+#define ETHER_HEAD_LEN      14
+
+#define PAR0_ECP_IRQ_NBR    4
+
+#define RX_DESC_BUF_SIZE   256
+#define NBR_OF_RX_DESC     (RX_BUF_SIZE / \
+			    RX_DESC_BUF_SIZE)
+
+/* Size of slave etrax boot image */
+#define ETRAX_PAR_BOOT_LENGTH 784
+
+static etrax_dma_descr *myNextRxDesc;  /* Points to the next descriptor to
+					  to be processed */
+static etrax_dma_descr *myLastRxDesc;  /* The last processed descriptor */
+static etrax_dma_descr *myPrevRxDesc;  /* The descriptor right before myNextRxDesc */
+
+static unsigned char RxBuf[RX_BUF_SIZE];
+
+static etrax_dma_descr RxDescList[NBR_OF_RX_DESC] __attribute__ ((aligned(4)));
+static etrax_dma_descr TxDescList[3] __attribute__ ((aligned(4)));
+                       /* host command, data, bogus ECP command */
+
+static struct sk_buff *tx_skb;
+
+/* Index to functions, as function prototypes. */
+
+static int etrax_ethernet_lpslave_init(struct net_device *dev);
+
+static int e100_open(struct net_device *dev);
+static int e100_set_mac_address(struct net_device *dev, void *addr);
+static int e100_send_packet(struct sk_buff *skb, struct net_device *dev);
+static void e100rx_interrupt(int irq, void *dev_id, struct pt_regs *regs);
+static void e100tx_interrupt(int irq, void *dev_id, struct pt_regs *regs);
+static void ecp_interrupt(int irq, void *dev_id, struct pt_regs *regs);
+static void e100_rx(struct net_device *dev);
+static int e100_close(struct net_device *dev);
+static struct net_device_stats *e100_get_stats(struct net_device *dev);
+static void set_multicast_list(struct net_device *dev);
+static void e100_hardware_send_packet(unsigned long hostcmd, char *buf, int length);
+static void update_rx_stats(struct net_device_stats *);
+static void update_tx_stats(struct net_device_stats *);
+static void e100_reset_tranceiver(void);
+
+static void boot_slave(unsigned char *code);
+
+#ifdef ETHDEBUG
+static void dump_parport_status(void);
+#endif
+
+#define tx_done(dev) (*R_DMA_CH0_CMD == 0)
+
+static unsigned long host_command;
+extern unsigned char e100lpslaveprog;
+
+/*
+ * This driver uses PAR0 to recevice data from slave ETRAX and PAR1 to boot
+ * and send data to slave ETRAX.
+ * Used ETRAX100 DMAchannels with corresponding IRQ:
+ * PAR0 RX : DMA3 - IRQ 19
+ * PAR1 TX : DMA4 - IRQ 20
+ * IRQ 4 is used to detect ECP commands from slave ETRAX
+ *
+ * NOTE! PAR0 and PAR1 shares DMA and IRQ numbers with SER2 and SER3 
+ */
+
+
+/*
+ * Check for a network adaptor of this type, and return '0' if one exists.
+ * If dev->base_addr == 0, probe all likely locations.
+ * If dev->base_addr == 1, always return failure.
+ * If dev->base_addr == 2, allocate space for the device and return success
+ * (detachable devices only).
+ */
+static int __init
+etrax_ethernet_lpslave_init(struct net_device *dev)
+{
+	int i;
+	int anOffset = 0;
+
+	printk("Etrax/100 lpslave ethernet driver v0.3, (c) 1999 Axis Communications AB\n");
+
+	dev->base_addr = 2;
+
+	printk("%s initialized\n", dev->name);
+
+	/* make Linux aware of the new hardware  */
+
+	if (!dev) {
+		printk(KERN_WARNING "%s: dev == NULL. Should this happen?\n",
+                       cardname);
+		dev = init_etherdev(dev, sizeof(struct net_local));
+		if (!dev)
+			panic("init_etherdev failed\n");
+	}
+
+	/* setup generic handlers and stuff in the dev struct */
+
+	ether_setup(dev);
+
+	/* make room for the local structure containing stats etc */
+
+	dev->priv = kmalloc(sizeof(struct net_local), GFP_KERNEL);
+	if (dev->priv == NULL)
+		return -ENOMEM;
+	memset(dev->priv, 0, sizeof(struct net_local));
+
+	/* now setup our etrax specific stuff */
+
+	dev->irq = DMA3_RX_IRQ_NBR; /* we really use DMATX as well... */
+        dev->dma = PAR0_RX_DMA_NBR;
+
+	/* fill in our handlers so the network layer can talk to us in the future */
+
+	dev->open               = e100_open;
+	dev->hard_start_xmit    = e100_send_packet;
+	dev->stop               = e100_close;
+	dev->get_stats          = e100_get_stats;
+	dev->set_multicast_list = set_multicast_list;
+	dev->set_mac_address    = e100_set_mac_address;
+        
+        /* set the default MAC address */
+
+	e100_set_mac_address(dev, &default_mac);
+
+	/* Initialise the list of Etrax DMA-descriptors */
+
+	/* Initialise receive descriptors */
+
+	for(i = 0; i < (NBR_OF_RX_DESC - 1); i++) {
+		RxDescList[i].ctrl   = 0;
+		RxDescList[i].sw_len = RX_DESC_BUF_SIZE;
+		RxDescList[i].next   = virt_to_phys(&RxDescList[i + 1]);
+		RxDescList[i].buf    = virt_to_phys(RxBuf + anOffset);
+		RxDescList[i].status = 0;
+		RxDescList[i].hw_len = 0;
+		anOffset += RX_DESC_BUF_SIZE;
+	}
+
+	RxDescList[i].ctrl   = d_eol;
+	RxDescList[i].sw_len = RX_DESC_BUF_SIZE;
+	RxDescList[i].next   = virt_to_phys(&RxDescList[0]);
+	RxDescList[i].buf    = virt_to_phys(RxBuf + anOffset);
+	RxDescList[i].status = 0;
+	RxDescList[i].hw_len = 0;
+
+	/* Initialise initial pointers */
+
+	myNextRxDesc = &RxDescList[0];
+	myLastRxDesc = &RxDescList[NBR_OF_RX_DESC - 1];
+	myPrevRxDesc = &RxDescList[NBR_OF_RX_DESC - 1];
+
+        /* setup some TX descriptor data */
+
+	TxDescList[0].sw_len = 4;
+	TxDescList[0].ctrl = 0;
+	TxDescList[0].buf = virt_to_phys(&host_command);
+	TxDescList[0].next = virt_to_phys(&TxDescList[1]);
+
+	return 0;
+}
+
+/* set MAC address of the interface. called from the core after a
+ * SIOCSIFADDR ioctl, and from the bootup above.
+ */
+
+static int
+e100_set_mac_address(struct net_device *dev, void *p)
+{
+	struct sockaddr *addr = p;
+	int i;
+
+	/* remember it */
+
+        memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
+
+	/* Write it to the hardware.
+	 * Note the way the address is wrapped:
+	 * *R_NETWORK_SA_0 = a0_0 | (a0_1 << 8) | (a0_2 << 16) | (a0_3 << 24);
+	 * *R_NETWORK_SA_1 = a0_4 | (a0_5 << 8);
+	 */
+
+        tx_skb = 0;
+	e100_hardware_send_packet(HOST_CMD_SETMAC, dev->dev_addr, 6);
+
+	/* show it in the log as well */
+
+	printk("%s: changed MAC to ", dev->name);
+
+	for (i = 0; i < 5; i++)
+		printk("%02X:", dev->dev_addr[i]);
+
+	printk("%02X\n", dev->dev_addr[i]);
+
+	return 0;
+}
+
+/*
+ * Open/initialize the board. This is called (in the current kernel)
+ * sometime after booting when the 'ifconfig' program is run.
+ *
+ * This routine should set everything up anew at each open, even
+ * registers that "should" only need to be set once at boot, so that
+ * there is non-reboot way to recover if something goes wrong.
+ */
+
+static int
+e100_open(struct net_device *dev)
+{
+	unsigned long flags;
+
+	/* configure the PAR0 (RX) and PAR1 (TX) ports
+	 *
+	 * perror is nAckReverse, which must be 1 at the TX side,
+         * and 0 at the RX side
+         *
+	 * select is XFlag, which must be 1 at both sides
+	 */
+#ifdef ETHDEBUG        
+        printk("Setting up PAR ports\n");
+#endif
+        *R_PAR0_CONFIG =
+          /* We do not have an external buffer, don't care */
+		IO_STATE(R_PAR0_CONFIG, ioe,     noninv)    |
+          /* Not connected, don't care */
+		IO_STATE(R_PAR0_CONFIG, iseli,   noninv)    |
+          /* iautofd is not inverted, noninv */
+		IO_STATE(R_PAR0_CONFIG, iautofd, noninv)    |
+          /* Not used in reverse direction, don't care */
+		IO_STATE(R_PAR0_CONFIG, istrb,   noninv)    |
+          /* Not connected, don't care /
+		IO_STATE(R_PAR0_CONFIG, iinit,   noninv)    |
+          /* perror is GND and reverse wants 0, noninv */
+		IO_STATE(R_PAR0_CONFIG, iperr,   noninv)    |
+          /* ack is not inverted, noninv */
+		IO_STATE(R_PAR0_CONFIG, iack,    noninv)    |
+          /* busy is not inverted, noninv */
+		IO_STATE(R_PAR0_CONFIG, ibusy,   noninv)    |
+          /* fault is not inverted, noninv */
+		IO_STATE(R_PAR0_CONFIG, ifault,  noninv)    |
+          /* select is Vcc and we want 1, noninv */
+		IO_STATE(R_PAR0_CONFIG, isel,    noninv)    |
+          /* We will run dma, enable */
+		IO_STATE(R_PAR0_CONFIG, dma, enable)        |
+          /* No run length encoding, disable */
+		IO_STATE(R_PAR0_CONFIG, rle_in, disable)    |
+          /* No run length encoding, disable */
+		IO_STATE(R_PAR0_CONFIG, rle_out, disable)   |
+          /* Enable parallel port */
+		IO_STATE(R_PAR0_CONFIG, enable, on)         |
+          /* Force mode regardless of pin status */
+		IO_STATE(R_PAR0_CONFIG, force, on)          |
+          /* We want ECP forward mode since PAR0 is RX */
+		IO_STATE(R_PAR0_CONFIG, mode, ecp_rev);       
+
+        *R_PAR1_CONFIG =
+          /* We do not have an external buffer, don't care */
+		IO_STATE(R_PAR1_CONFIG, ioe,     noninv)    |
+          
+          /* Not connected, don't care */
+		IO_STATE(R_PAR1_CONFIG, iseli,   noninv)    |
+          
+          /* HostAck must indicate data cycle, noninv */
+		IO_STATE(R_PAR1_CONFIG, iautofd, noninv)    |
+          
+          /* HostClk has no external inverter, noninv */
+		IO_STATE(R_PAR1_CONFIG, istrb,   noninv)    |
+          
+          /* Not connected, don't care */
+		IO_STATE(R_PAR1_CONFIG, iinit,   noninv)    |
+          
+          /* nAckReverse must be 1 in forward mode but is grounded, inv */ 
+		IO_STATE(R_PAR1_CONFIG, iperr,   inv)       |
+          
+          /* PeriphClk must be 1 in forward mode, noninv */
+		IO_STATE(R_PAR1_CONFIG, iack,    noninv)    |
+          
+          /* PeriphAck has no external inverter, noninv */
+		IO_STATE(R_PAR1_CONFIG, ibusy,   noninv)    |
+          
+          /* nPerihpRequest has no external inverter, noniv */
+		IO_STATE(R_PAR1_CONFIG, ifault,  noninv)    |
+          
+          /* Select is VCC and we want 1, noninv */
+		IO_STATE(R_PAR1_CONFIG, isel,    noninv)    |
+          
+          /* No EPP mode, disable */
+                IO_STATE(R_PAR1_CONFIG, ext_mode, disable)  |
+          
+          /* We will run dma, enable */
+                IO_STATE(R_PAR1_CONFIG, dma, enable)        |
+          
+          /* No run length encoding, disable */
+		IO_STATE(R_PAR1_CONFIG, rle_in, disable)    |
+          
+          /* No run length encoding, disable */
+		IO_STATE(R_PAR1_CONFIG, rle_out, disable)   |
+          
+          /* Enable parallel port */
+		IO_STATE(R_PAR1_CONFIG, enable, on)         |
+          
+          /* Force mode regardless of pin status */
+		IO_STATE(R_PAR1_CONFIG, force, on)          |
+          
+          /* We want ECP forward mode since PAR1 is TX */
+	 	IO_STATE(R_PAR1_CONFIG, mode, ecp_fwd);        
+
+        /* Setup time of value * 160 + 20 ns == 180 ns below */
+        *R_PAR1_DELAY = IO_FIELD(R_PAR1_DELAY, setup, 1);  
+
+        *R_PAR1_CTRL = 0;
+
+        while ((((*R_PAR1_STATUS)&0xE000) >> 13) != 5); /* Wait for ECP_FWD mode */
+#ifdef ETHDEBUG
+        dump_parport_status();
+#endif
+        
+        /* make sure ECP irq is acked when we enable it below */
+
+	(void)*R_PAR0_STATUS_DATA;
+	(void)*R_PAR1_STATUS_DATA;
+
+	/* Reset and wait for the DMA channels */
+
+        RESET_DMA(4); /* PAR1_TX_DMA_NBR */
+	RESET_DMA(3); /* PAR0_RX_DMA_NBR */
+	WAIT_DMA(4);  
+	WAIT_DMA(3);
+        
+        /* boot the slave Etrax, by sending code on PAR1.
+	 * do this before we start up the IRQ handlers and stuff,
+	 * beacuse we simply poll for completion in boot_slave.
+	 */
+        
+	boot_slave(&e100lpslaveprog);
+
+	/* allocate the irq corresponding to the receiving DMA */
+
+	if (request_irq(DMA3_RX_IRQ_NBR, e100rx_interrupt, 0,
+			cardname, (void *)dev)) {
+          printk("Failed to allocate DMA3_RX_IRQ_NBR\n");
+		goto grace_exit;
+	}
+
+	/* allocate the irq corresponding to the transmitting DMA */
+
+	if (request_irq(DMA4_TX_IRQ_NBR, e100tx_interrupt, 0,
+			cardname, (void *)dev)) {
+          printk("Failed to allocate DMA4_TX_IRQ_NBR\n");
+          goto grace_exit;
+	}
+        
+        /* allocate the irq used for detecting ECP commands on the RX port (PAR0) */
+
+	if (request_irq(PAR0_ECP_IRQ_NBR, ecp_interrupt, 0,
+			cardname, (void *)dev)) {
+          printk("Failed to allocate PAR0_ECP_IRQ_NBR\n");
+          grace_exit:          
+                free_irq(PAR0_ECP_IRQ_NBR, (void *)dev);
+                free_irq(DMA4_TX_IRQ_NBR, (void *)dev);
+		free_irq(DMA3_RX_IRQ_NBR, (void *)dev);
+                
+		return -EAGAIN;
+	}
+
+#if 0
+        /* We are not allocating DMA since DMA4 is reserved for 'cascading'
+         * and will always fail with the current dma.c
+         */
+        
+	/*
+	 * Always allocate the DMA channels after the IRQ,
+	 * and clean up on failure.
+	 */
+
+	if(request_dma(PAR0_RX_DMA_NBR, cardname)) {
+          printk("Failed to allocate PAR0_RX_DMA_NBR\n");
+		goto grace_exit;
+	}
+
+	if(request_dma(PAR1_TX_DMA_NBR, cardname)) {
+          printk("Failed to allocate PAR1_TX_DMA_NBR\n");
+	grace_exit:
+		/* this will cause some 'trying to free free irq' but what the heck... */
+
+		free_dma(PAR1_TX_DMA_NBR);
+                free_dma(PAR0_RX_DMA_NBR);
+                free_irq(PAR0_ECP_IRQ_NBR, (void *)dev);
+                free_irq(DMA4_TX_IRQ_NBR, (void *)dev);
+		free_irq(DMA3_RX_IRQ_NBR, (void *)dev);
+		
+		return -EAGAIN;
+	}
+#endif
+        
+#ifdef ETHDEBUG
+        printk("Par port IRQ and DMA allocated\n");
+#endif
+	save_flags(flags);
+	cli();
+
+	/* enable the irq's for PAR0/1 DMA */
+
+	*R_IRQ_MASK2_SET =
+		IO_STATE(R_IRQ_MASK2_SET, dma3_eop, set) |
+		IO_STATE(R_IRQ_MASK2_SET, dma4_descr, set);
+
+        *R_IRQ_MASK0_SET =
+		IO_STATE(R_IRQ_MASK0_SET, par0_ecp_cmd, set);
+
+	tx_skb = 0;
+
+	/* make sure the irqs are cleared */
+
+	*R_DMA_CH3_CLR_INTR = IO_STATE(R_DMA_CH3_CLR_INTR, clr_eop, do);
+	*R_DMA_CH4_CLR_INTR = IO_STATE(R_DMA_CH4_CLR_INTR, clr_descr, do);
+
+        /* Write the MAC address to the slave HW */
+	udelay(5000);
+	e100_hardware_send_packet(HOST_CMD_SETMAC, dev->dev_addr, 6);
+        
+	/* make sure the rec and transmit error counters are cleared */
+
+	(void)*R_REC_COUNTERS;  /* dummy read */
+	(void)*R_TR_COUNTERS;   /* dummy read */
+
+	/* start the receiving DMA channel so we can receive packets from now on */
+
+	*R_DMA_CH3_FIRST = virt_to_phys(myNextRxDesc);
+	*R_DMA_CH3_CMD = IO_STATE(R_DMA_CH3_CMD, cmd, start);
+
+	restore_flags(flags);
+	
+	/* We are now ready to accept transmit requeusts from
+	 * the queueing layer of the networking.
+	 */
+#ifdef ETHDEBUG
+        printk("Starting slave network transmit queue\n");
+#endif
+	netif_start_queue(dev);
+
+	return 0;
+}
+
+static void 
+e100_reset_tranceiver(void)
+{
+  /* To do: Reboot and setup slave Etrax */
+}
+
+/* Called by upper layers if they decide it took too long to complete
+ * sending a packet - we need to reset and stuff.
+ */
+
+static void
+e100_tx_timeout(struct net_device *dev)
+{
+	struct net_local *np = (struct net_local *)dev->priv;
+
+	printk(KERN_WARNING "%s: transmit timed out, %s?\n", dev->name,
+	       tx_done(dev) ? "IRQ problem" : "network cable problem");
+	
+	/* remember we got an error */
+	
+	np->stats.tx_errors++; 
+	
+	/* reset the TX DMA in case it has hung on something */
+	
+	RESET_DMA(4);
+	WAIT_DMA(4);
+	
+	/* Reset the tranceiver. */
+	
+	e100_reset_tranceiver();
+	
+	/* and get rid of the packet that never got an interrupt */
+	
+	dev_kfree_skb(tx_skb);
+	tx_skb = 0;
+	
+	/* tell the upper layers we're ok again */
+	
+	netif_wake_queue(dev);
+}
+
+
+/* This will only be invoked if the driver is _not_ in XOFF state.
+ * What this means is that we need not check it, and that this
+ * invariant will hold if we make sure that the netif_*_queue()
+ * calls are done at the proper times.
+ */
+
+static int
+e100_send_packet(struct sk_buff *skb, struct net_device *dev)
+{
+	struct net_local *np = (struct net_local *)dev->priv;
+	int length = ETH_ZLEN < skb->len ? skb->len : ETH_ZLEN;
+	unsigned char *buf = skb->data;
+	
+#ifdef ETHDEBUG
+        unsigned char *temp_data_ptr = buf;
+        int i;
+        
+	printk("Sending a packet of length %d:\n", length);
+	/* dump the first bytes in the packet */
+	for(i = 0; i < 8; i++) {
+		printk("%d: %.2x %.2x %.2x %.2x %.2x %.2x %.2x %.2x\n", i * 8,
+		       temp_data_ptr[0],temp_data_ptr[1],temp_data_ptr[2],
+                       temp_data_ptr[3],temp_data_ptr[4],temp_data_ptr[5],
+                       temp_data_ptr[6],temp_data_ptr[7]);
+		temp_data_ptr += 8;
+	}
+#endif
+	spin_lock_irq(&np->lock);  /* protect from tx_interrupt */
+
+	tx_skb = skb; /* remember it so we can free it in the tx irq handler later */
+	dev->trans_start = jiffies;
+	
+	e100_hardware_send_packet(HOST_CMD_SENDPACK, buf, length);
+
+	/* this simple TX driver has only one send-descriptor so we're full
+	 * directly. If this had a send-ring instead, we would only do this if
+	 * the ring got full.
+	 */
+
+	netif_stop_queue(dev);
+        
+	spin_unlock_irq(&np->lock);
+
+	return 0;
+}
+
+/*
+ * The typical workload of the driver:
+ *   Handle the network interface interrupts.
+ */
+
+static void
+e100rx_interrupt(int irq, void *dev_id, struct pt_regs * regs)
+{
+	struct net_device *dev = (struct net_device *)dev_id;
+	unsigned long irqbits = *R_IRQ_MASK2_RD;
+        
+	if(irqbits & IO_STATE(R_IRQ_MASK2_RD, dma3_eop, active)) {
+
+		/* acknowledge the eop interrupt */
+
+		*R_DMA_CH3_CLR_INTR = IO_STATE(R_DMA_CH3_CLR_INTR, clr_eop, do);
+
+		/* check if one or more complete packets were indeed received */
+
+		while(*R_DMA_CH3_FIRST != virt_to_phys(myNextRxDesc)) {
+			/* Take out the buffer and give it to the OS, then
+			 * allocate a new buffer to put a packet in.
+			 */
+			e100_rx(dev);
+			((struct net_local *)dev->priv)->stats.rx_packets++;
+			/* restart/continue on the channel, for safety */
+			*R_DMA_CH3_CMD = IO_STATE(R_DMA_CH3_CMD, cmd, restart);
+			/* clear dma channel 3 eop/descr irq bits */
+			*R_DMA_CH3_CLR_INTR =
+				IO_STATE(R_DMA_CH3_CLR_INTR, clr_eop, do) |
+				IO_STATE(R_DMA_CH3_CLR_INTR, clr_descr, do);
+			
+			/* now, we might have gotten another packet
+			   so we have to loop back and check if so */
+		}
+	}
+}
+
+/* the transmit dma channel interrupt
+ *
+ * this is supposed to free the skbuff which was pending during transmission,
+ * and inform the kernel that we can send one more buffer
+ */
+
+static void
+e100tx_interrupt(int irq, void *dev_id, struct pt_regs * regs)
+{
+	struct net_device *dev = (struct net_device *)dev_id;
+	unsigned long irqbits = *R_IRQ_MASK2_RD;
+	struct net_local *np = (struct net_local *)dev->priv;
+
+#ifdef ETHDEBUG
+        printk("We got tx interrupt\n");
+#endif
+	/* check for a dma4_eop interrupt */
+	if(irqbits & IO_STATE(R_IRQ_MASK2_RD, dma4_descr, active)) {
+		/* This protects us from concurrent execution of
+		 * our dev->hard_start_xmit function above.
+		 */
+
+		spin_lock(&np->lock);
+		
+		/* acknowledge the eop interrupt */
+                
+		*R_DMA_CH4_CLR_INTR = IO_STATE(R_DMA_CH4_CLR_INTR, clr_descr, do);
+
+                /* skip *R_DMA_CH4_FIRST == 0 test since we use d_wait... */
+		if(tx_skb) {
+
+			np->stats.tx_bytes += tx_skb->len;
+			np->stats.tx_packets++;
+			/* dma is ready with the transmission of the data in tx_skb, so now we can release the skb memory */
+			dev_kfree_skb_irq(tx_skb);
+			tx_skb = 0;
+			netif_wake_queue(dev);
+		} else {
+			printk(KERN_WARNING "%s: tx weird interrupt\n",
+                               cardname);
+		}
+
+		spin_unlock(&np->lock);
+	}
+}
+
+static void
+ecp_interrupt(int irq, void *dev_id, struct pt_regs * regs)
+{
+	struct net_device *dev = (struct net_device *)dev_id;
+	struct net_local *lp = (struct net_local *)dev->priv;
+	unsigned long temp, irqbits = *R_IRQ_MASK0_RD;
+
+        /* check for ecp irq */
+	if(irqbits & IO_MASK(R_IRQ_MASK0_RD, par0_ecp_cmd)) { 
+		/* acknowledge by reading the bit */
+		temp = *R_PAR0_STATUS_DATA;
+		/* force an EOP on the incoming channel, so we'll get an rx interrupt */
+		*R_SET_EOP = IO_STATE(R_SET_EOP, ch3_eop, set);
+	}
+}
+
+/* We have a good packet(s), get it/them out of the buffers. */
+static void
+e100_rx(struct net_device *dev)
+{
+	struct sk_buff *skb;
+	int length=0;
+	int i;
+	struct net_local *np = (struct net_local *)dev->priv;
+	struct etrax_dma_descr *mySaveRxDesc = myNextRxDesc;
+	unsigned char *skb_data_ptr;
+
+	/* If the packet is broken down in many small packages then merge
+	 * count how much space we will need to alloc with skb_alloc() for
+	 * it to fit.
+	 */
+
+	while (!(myNextRxDesc->status & d_eop)) {
+		length += myNextRxDesc->sw_len; /* use sw_len for the first descs */
+		myNextRxDesc->status = 0;
+		myNextRxDesc = phys_to_virt(myNextRxDesc->next);
+	}
+
+	length += myNextRxDesc->hw_len; /* use hw_len for the last descr */
+
+#ifdef ETHDEBUG
+	printk("Got a packet of length %d:\n", length);
+	/* dump the first bytes in the packet */
+	skb_data_ptr = (unsigned char *)phys_to_virt(mySaveRxDesc->buf);
+	for(i = 0; i < 8; i++) {
+		printk("%d: %.2x %.2x %.2x %.2x %.2x %.2x %.2x %.2x\n", i * 8,
+		       skb_data_ptr[0],skb_data_ptr[1],skb_data_ptr[2],skb_data_ptr[3],
+		       skb_data_ptr[4],skb_data_ptr[5],skb_data_ptr[6],skb_data_ptr[7]);
+		skb_data_ptr += 8;
+	}
+#endif
+
+	skb = dev_alloc_skb(length - ETHER_HEAD_LEN);
+	if (!skb) {
+		np->stats.rx_errors++;
+		printk(KERN_NOTICE "%s: Memory squeeze, dropping packet.\n",
+		       dev->name);
+		return;
+	}
+
+	skb_put(skb, length - ETHER_HEAD_LEN);        /* allocate room for the packet body */
+	skb_data_ptr = skb_push(skb, ETHER_HEAD_LEN); /* allocate room for the header */
+
+#ifdef ETHDEBUG
+	printk("head = 0x%x, data = 0x%x, tail = 0x%x, end = 0x%x\n",
+	       skb->head, skb->data, skb->tail, skb->end);
+	printk("copying packet to 0x%x.\n", skb_data_ptr);
+#endif
+
+	/* this loop can be made using max two memcpy's if optimized */
+
+	while(mySaveRxDesc != myNextRxDesc) {
+		memcpy(skb_data_ptr, phys_to_virt(mySaveRxDesc->buf),
+		       mySaveRxDesc->sw_len);
+		skb_data_ptr += mySaveRxDesc->sw_len;
+		mySaveRxDesc = phys_to_virt(mySaveRxDesc->next);
+	}
+
+	memcpy(skb_data_ptr, phys_to_virt(mySaveRxDesc->buf),
+	       mySaveRxDesc->hw_len);
+
+	skb->dev = dev;
+	skb->protocol = eth_type_trans(skb, dev);
+
+	/* Send the packet to the upper layers */
+
+	netif_rx(skb);
+
+	/* Prepare for next packet */
+
+	myNextRxDesc->status = 0;
+	myPrevRxDesc = myNextRxDesc;
+	myNextRxDesc = phys_to_virt(myNextRxDesc->next);
+
+	myPrevRxDesc->ctrl |= d_eol;
+	myLastRxDesc->ctrl &= ~d_eol;
+	myLastRxDesc = myPrevRxDesc;
+
+	return;
+}
+
+/* The inverse routine to net_open(). */
+static int
+e100_close(struct net_device *dev)
+{
+	struct net_local *np = (struct net_local *)dev->priv;
+
+	printk("Closing %s.\n", dev->name);
+
+	netif_stop_queue(dev);
+
+	*R_IRQ_MASK0_CLR = IO_STATE(R_IRQ_MASK0_CLR, par0_ecp_cmd, clr);
+
+	*R_IRQ_MASK2_CLR =
+		IO_STATE(R_IRQ_MASK2_CLR, dma3_eop, clr) |
+		IO_STATE(R_IRQ_MASK2_CLR, dma4_descr, clr);	
+
+	/* Stop the receiver and the transmitter */
+
+	RESET_DMA(3);
+	RESET_DMA(4);
+
+	/* Flush the Tx and disable Rx here. */
+
+	free_irq(DMA3_RX_IRQ_NBR, (void *)dev);
+	free_irq(DMA4_TX_IRQ_NBR, (void *)dev);
+	free_irq(PAR0_ECP_IRQ_NBR, (void *)dev);
+
+	free_dma(PAR1_TX_DMA_NBR);
+	free_dma(PAR0_RX_DMA_NBR);
+
+	/* Update the statistics here. */
+
+	update_rx_stats(&np->stats);
+	update_tx_stats(&np->stats);
+
+	return 0;
+}
+
+static void
+update_rx_stats(struct net_device_stats *es)
+{
+	unsigned long r = *R_REC_COUNTERS;
+	/* update stats relevant to reception errors */
+	es->rx_fifo_errors += r >> 24;            /* fifo overrun */
+	es->rx_crc_errors += r & 0xff;            /* crc error */
+	es->rx_frame_errors += (r >> 8) & 0xff;   /* alignment error */
+	es->rx_length_errors += (r >> 16) & 0xff; /* oversized frames */
+}
+
+static void
+update_tx_stats(struct net_device_stats *es)
+{
+	unsigned long r = *R_TR_COUNTERS;
+	/* update stats relevant to transmission errors */
+	es->collisions += (r & 0xff) + ((r >> 8) & 0xff); /* single_col + multiple_col */
+	es->tx_errors += (r >> 24) & 0xff; /* deferred transmit frames */
+}
+
+/*
+ * Get the current statistics.
+ * This may be called with the card open or closed.
+ */
+static struct net_device_stats *
+e100_get_stats(struct net_device *dev)
+{
+	struct net_local *lp = (struct net_local *)dev->priv;
+
+	update_rx_stats(&lp->stats);
+	update_tx_stats(&lp->stats);
+
+	return &lp->stats;
+}
+
+/*
+ * Set or clear the multicast filter for this adaptor.
+ * num_addrs == -1	Promiscuous mode, receive all packets
+ * num_addrs == 0	Normal mode, clear multicast list
+ * num_addrs > 0	Multicast mode, receive normal and MC packets,
+ *			and do best-effort filtering.
+ */
+static void
+set_multicast_list(struct net_device *dev)
+{
+  /* To do */
+}
+
+void
+e100_hardware_send_packet(unsigned long hostcmd, char *buf, int length)
+{
+  static char bogus_ecp[] = { 42, 42 };
+  int i;
+  
+
+#ifdef ETHDEBUG
+	printk("e100 send pack, buf 0x%x len %d\n", buf, length);
+#endif
+        
+        host_command = hostcmd;
+
+	/* Configure the tx dma descriptor. Desc 0 is already configured.*/
+
+        TxDescList[1].sw_len = length;
+	/* bug workaround - etrax100 needs d_wait on the descriptor _before_
+	 * a descriptor containing an ECP command
+	 */
+	TxDescList[1].ctrl = d_wait;
+	TxDescList[1].buf = virt_to_phys(buf);
+	TxDescList[1].next = virt_to_phys(&TxDescList[2]);
+
+        /* append the ecp dummy descriptor - its only purpose is to 
+	 * make the receiver generate an irq due to the ecp command
+	 * so the receiver knows where packets end
+	 */
+
+	TxDescList[2].sw_len = 1;
+	TxDescList[2].ctrl = d_ecp | d_eol | d_int;
+	TxDescList[2].buf = virt_to_phys(bogus_ecp);
+        
+
+	/* setup the dma channel and start it */
+
+        *R_DMA_CH4_FIRST = virt_to_phys(TxDescList);
+	*R_DMA_CH4_CMD = IO_STATE(R_DMA_CH4_CMD, cmd, start);
+        
+#ifdef ETHDEBUG         
+         printk("done\n");
+#endif
+}
+
+/* send a chunk of code to the slave chip to boot it. */
+
+static void
+boot_slave(unsigned char *code)
+{
+  int i;
+
+#ifdef ETHDEBUG
+	printk("  booting slave ETRAX...\n");
+#endif
+        *R_PORT_PB_DATA = 0x7F; /* Reset slave */
+        udelay(15); /* Time enough to reset WAN tranciever */
+        *R_PORT_PB_DATA = 0xFF; /* Reset slave */
+
+	/* configure the tx dma data descriptor */
+
+	TxDescList[1].sw_len = ETRAX_PAR_BOOT_LENGTH;
+	TxDescList[1].ctrl = d_eol | d_int; 
+                                                      
+	TxDescList[1].buf = virt_to_phys(code);
+        /*TxDescList[1].buf = code;*/
+	TxDescList[1].next = 0;
+        
+        /* setup the dma channel and start it */
+ 	*R_DMA_CH4_FIRST = virt_to_phys(&TxDescList[1]);
+	*R_DMA_CH4_CMD = IO_STATE(R_DMA_CH4_CMD, cmd, start);
+
+	/* wait for completion */
+	while(!(*R_IRQ_READ2 & IO_MASK(R_IRQ_READ2, dma4_descr)));
+
+	/* ack the irq */
+
+	*R_DMA_CH4_CLR_INTR = IO_STATE(R_DMA_CH4_CLR_INTR, clr_descr, do);
+
+#if 0
+        /* manual transfer of boot code - requires dma turned off */
+        for (i=0; i<ETRAX_PAR_BOOT_LENGTH; i++)
+        {
+          printk("  sending byte: %u value: %x\n",i,code[i]);
+          while (((*R_PAR1_STATUS)&0x02) == 0); /* Wait while tr_rdy is busy*/
+          *R_PAR1_CTRL_DATA = code[i];
+        }
+#endif
+
+#ifdef ETHDEBUG
+	printk("  done\n");
+#endif
+}
+
+#ifdef ETHDEBUG
+/* debug code to check the current status of PAR1 */
+static void
+dump_parport_status(void)
+{
+  unsigned long temp;
+  
+  printk("Parport1 status:\n");
+  
+  temp = (*R_PAR1_STATUS)&0xE000;
+  temp = temp >> 13;
+  printk("Reg mode: %u (ecp_fwd(5), ecp_rev(6))\n", temp);
+  
+  temp = (*R_PAR1_STATUS)&0x1000;
+  temp = temp >> 12;
+  printk("Reg perr: %u (ecp_rev(0))\n", temp);
+  
+  temp = (*R_PAR1_STATUS)&0x0800;
+  temp = temp >> 11;
+  printk("Reg ack: %u (inactive (1), active (0))\n", temp);
+  
+  temp = (*R_PAR1_STATUS)&0x0400;
+  temp = temp >> 10;
+  printk("Reg busy: %u (inactive (0), active (1))\n", temp);
+  
+  temp = (*R_PAR1_STATUS)&0x0200;
+  temp = temp >> 9;
+  printk("Reg fault: %u (inactive (1), active (0))\n", temp);
+  
+  temp = (*R_PAR1_STATUS)&0x0100;
+  temp = temp >> 8;
+  printk("Reg sel: %u (inactive (0), active (1), xflag(1))\n", temp);
+
+  temp = (*R_PAR1_STATUS)&0x02;
+  temp = temp >> 1;
+  printk("Reg tr_rdy: %u (busy (0), ready (1))\n", temp);
+
+}
+#endif /* ETHDEBUG */
+
+static struct net_device dev_etrax_slave_ethernet;
+
+static int
+etrax_init_module(void)
+{
+	struct net_device *d = &dev_etrax_slave_ethernet;
+
+	d->init = etrax_ethernet_lpslave_init;
+
+	if(register_netdev(d) == 0)
+		return 0;
+	else
+		return -ENODEV;
+}
+
+module_init(etrax_init_module);

FUNET's LINUX-ADM group, linux-adm@nic.funet.fi
TCL-scripts by Sam Shen (who was at: slshen@lbl.gov)