patch-2.4.3 linux/drivers/net/wan/dscc4.c

Next file: linux/drivers/net/wan/hd64570.h
Previous file: linux/drivers/net/wan/dlci.c
Back to the patch index
Back to the overall index

diff -u --recursive --new-file v2.4.2/linux/drivers/net/wan/dscc4.c linux/drivers/net/wan/dscc4.c
@@ -0,0 +1,1769 @@
+/*
+ * drivers/net/wan/dscc4/dscc4_main.c: a DSCC4 HDLC driver for Linux
+ *
+ * This software may be used and distributed according to the terms of the 
+ * GNU General Public License. 
+ *
+ * The author may be reached as romieu@cogenit.fr.
+ * Specific bug reports/asian food will be welcome.
+ *
+ * Special thanks to the nice people at CS-Telecom for the hardware and the
+ * access to the test/measure tools.
+ *
+ *
+ *                             Theory of Operation
+ *
+ * I. Board Compatibility
+ *
+ * This device driver is designed for the Siemens PEB20534 4 ports serial
+ * controller as found on Etinc PCISYNC cards. The documentation for the 
+ * chipset is available at http://www.infineon.com:
+ * - Data Sheet "DSCC4, DMA Supported Serial Communication Controller with
+ * 4 Channels, PEB 20534 Version 2.1, PEF 20534 Version 2.1";
+ * - Application Hint "Management of DSCC4 on-chip FIFO resources".
+ * Jens David has built an adapter based on the same chipset. Take a look
+ * at http://www.afthd.tu-darmstadt.de/~dg1kjd/pciscc4 for a specific
+ * driver.
+ * Sample code (2 revisions) is available at Infineon.
+ *
+ * II. Board-specific settings
+ *
+ * Pcisync can transmit some clock signal to the outside world on the
+ * *first two* ports provided you put a quartz and a line driver on it and
+ * remove the jumpers. The operation is described on Etinc web site. If you
+ * go DCE on these ports, don't forget to use an adequate cable.
+ *
+ * Sharing of the PCI interrupt line for this board is possible.
+ *
+ * III. Driver operation
+ *
+ * The rx/tx operations are based on a linked list of descriptor. I haven't
+ * tried the start/stop descriptor method as this one looks like the cheapest
+ * in terms of PCI manipulation.
+ *
+ * Tx direction
+ * Once the data section of the current descriptor processed, the next linked
+ * descriptor is loaded if the HOLD bit isn't set in the current descriptor.
+ * If HOLD is met, the transmission is stopped until the host unsets it and
+ * signals the change via TxPOLL.
+ * When the tx ring is full, the xmit routine issues a call to netdev_stop.
+ * The device is supposed to be enabled again during an ALLS irq (we could
+ * use HI but as it's easy to loose events, it's fscked).
+ *
+ * Rx direction
+ * The received frames aren't supposed to span over multiple receiving areas.
+ * I may implement it some day but it isn't the highest ranked item.
+ *
+ * IV. Notes
+ * The chipset is buggy. Typically, under some specific load patterns (I
+ * wouldn't call them "high"), the irq queues and the descriptors look like
+ * some event has been lost. Even assuming some fancy PCI feature, it won't 
+ * explain the reproductible missing "C" bit in the descriptors. Faking an 
+ * irq in the periodic timer isn't really elegant but at least it seems 
+ * reliable.
+ * The current error (XDU, RFO) recovery code is untested.
+ * So far, RDO takes his RX channel down and the right sequence to enable it
+ * again is still a mistery. If RDO happens, plan a reboot. More details
+ * in the code (NB: as this happens, TX still works).
+ * Don't mess the cables during operation, especially on DTE ports. I don't
+ * suggest it for DCE either but at least one can get some messages instead
+ * of a complete instant freeze.
+ * Tests are done on Rev. 20 of the silicium. The RDO handling changes with
+ * the documentation/chipset releases. An on-line errata would be welcome.
+ *
+ * TODO:
+ * - some trivial error lurk,
+ * - the stats are fscked,
+ * - use polling at high irq/s,
+ * - performance analysis,
+ * - endianness.
+ *
+ */
+
+#include <linux/version.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/errno.h>
+#include <linux/ioport.h>
+#include <linux/pci.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+
+#include <asm/system.h>
+#include <asm/cache.h>
+#include <asm/byteorder.h>
+#include <asm/uaccess.h>
+#include <asm/io.h>
+#include <asm/irq.h>
+
+#include <linux/init.h>
+#include <linux/string.h>
+
+#include <linux/if_arp.h>
+#include <linux/netdevice.h>
+#include <linux/skbuff.h>
+#include <linux/delay.h>
+#include <net/syncppp.h>
+#include <linux/hdlc.h>
+
+/* Version */
+static const char * version = "$Id: dscc4.c,v 1.130 2001/02/25 15:27:34 romieu Exp $\n";
+static int debug;
+
+
+/* Module parameters */
+MODULE_AUTHOR("Maintainer: Francois Romieu <romieu@cogenit.fr>");
+MODULE_DESCRIPTION("Siemens PEB20534 PCI Controller");
+MODULE_PARM(debug,"i");
+
+/* Structures */
+struct TxFD {
+	u32 state;
+	u32 next;
+	u32 data;
+	u32 complete;
+	u32 jiffies; /* more hack to come :o) */
+};
+
+struct RxFD {
+	u32 state1;
+	u32 next;
+	u32 data;
+	u32 state2;
+	u32 end;
+};
+
+#define DEBUG
+#define DEBUG_PARANOID
+#define TX_RING_SIZE    32
+#define RX_RING_SIZE    32
+#define IRQ_RING_SIZE   64 /* Keep it A multiple of 32 */
+#define TX_TIMEOUT      (HZ/10)
+#define BRR_DIVIDER_MAX 64*0x00008000
+#define dev_per_card	4
+
+#define SOURCE_ID(flags) ((flags >> 28 ) & 0x03)
+#define TO_SIZE(state) ((state >> 16) & 0x1fff)
+#define TO_STATE(len) cpu_to_le32((len & TxSizeMax) << 16)
+#define RX_MAX(len) ((((len) >> 5) + 1) << 5)
+#define SCC_REG_START(id) SCC_START+(id)*SCC_OFFSET
+
+#undef DEBUG
+
+struct dscc4_pci_priv {
+        u32 *iqcfg;
+        int cfg_cur;
+        spinlock_t lock;
+        struct pci_dev *pdev;
+
+        struct net_device *root;
+        dma_addr_t iqcfg_dma;
+	u32 xtal_hz;
+};
+
+struct dscc4_dev_priv {
+        struct sk_buff *rx_skbuff[RX_RING_SIZE];
+        struct sk_buff *tx_skbuff[TX_RING_SIZE];
+
+        struct RxFD *rx_fd;
+        struct TxFD *tx_fd;
+        u32 *iqrx;
+        u32 *iqtx;
+
+        u32 rx_current;
+        u32 tx_current;
+        u32 iqrx_current;
+        u32 iqtx_current;
+
+        u32 tx_dirty;
+	int bad_tx_frame;
+	int bad_rx_frame;
+	int rx_needs_refill;
+
+        dma_addr_t tx_fd_dma;
+        dma_addr_t rx_fd_dma;
+        dma_addr_t iqtx_dma;
+        dma_addr_t iqrx_dma;
+
+        struct net_device_stats stats;
+	struct timer_list timer;
+
+        struct dscc4_pci_priv *pci_priv;
+        spinlock_t lock;
+
+        int dev_id;
+	u32 flags;
+	u32 timer_help;
+	u32 hi_expected;
+
+	struct hdlc_device_struct hdlc;
+	int usecount;
+};
+
+/* GLOBAL registers definitions */
+#define GCMDR   0x00
+#define GSTAR   0x04
+#define GMODE   0x08
+#define IQLENR0 0x0C
+#define IQLENR1 0x10
+#define IQRX0   0x14
+#define IQTX0   0x24
+#define IQCFG   0x3c
+#define FIFOCR1 0x44
+#define FIFOCR2 0x48
+#define FIFOCR3 0x4c
+#define FIFOCR4 0x34
+#define CH0CFG  0x50
+#define CH0BRDA 0x54
+#define CH0BTDA 0x58
+
+/* SCC registers definitions */
+#define SCC_START	0x0100
+#define SCC_OFFSET      0x80
+#define CMDR    0x00
+#define STAR    0x04
+#define CCR0    0x08
+#define CCR1    0x0c
+#define CCR2    0x10
+#define BRR     0x2C
+#define RLCR    0x40
+#define IMR     0x54
+#define ISR     0x58
+
+/* Bit masks */
+#define IntRxScc0       0x10000000
+#define IntTxScc0       0x01000000
+
+#define TxPollCmd	0x00000400
+#define RxActivate	0x08000000
+#define MTFi		0x04000000
+#define Rdr		0x00400000
+#define Rdt		0x00200000
+#define Idr		0x00100000
+#define Idt		0x00080000
+#define TxSccRes       0x01000000
+#define RxSccRes       0x00010000
+#define TxSizeMax	0x1ffc
+#define RxSizeMax	0x1ffc
+
+#define Ccr0ClockMask	0x0000003f
+#define Ccr1LoopMask	0x00000200
+#define BrrExpMask	0x00000f00
+#define BrrMultMask	0x0000003f
+#define EncodingMask	0x00700000
+#define Hold		0x40000000
+#define SccBusy		0x10000000
+#define FrameOk		(FrameVfr | FrameCrc)
+#define FrameVfr	0x80
+#define FrameRdo	0x40
+#define FrameCrc	0x20
+#define FrameAborted	0x00000200
+#define FrameEnd	0x80000000
+#define DataComplete	0x40000000
+#define LengthCheck	0x00008000
+#define SccEvt		0x02000000
+#define NoAck		0x00000200
+#define Action		0x00000001
+#define HiDesc		0x20000000
+
+/* SCC events */
+#define RxEvt		0xf0000000
+#define TxEvt		0x0f000000
+#define Alls		0x00040000
+#define Xdu		0x00010000
+#define Xmr		0x00002000
+#define Xpr		0x00001000
+#define Rdo		0x00000080
+#define Rfs		0x00000040
+#define Rfo		0x00000002
+#define Flex		0x00000001
+
+/* DMA core events */
+#define Cfg		0x00200000
+#define Hi		0x00040000
+#define Fi		0x00020000
+#define Err		0x00010000
+#define Arf		0x00000002
+#define ArAck		0x00000001
+
+/* Misc */
+#define NeedIDR		0x00000001
+#define NeedIDT		0x00000002
+#define RdoSet		0x00000004
+
+/* Functions prototypes */
+static __inline__ void dscc4_rx_irq(struct dscc4_pci_priv *, struct net_device *);
+static __inline__ void dscc4_tx_irq(struct dscc4_pci_priv *, struct net_device *);
+static int dscc4_found1(struct pci_dev *, unsigned long ioaddr);
+static int dscc4_init_one(struct pci_dev *, const struct pci_device_id *ent);
+static int dscc4_open(struct net_device *);
+static int dscc4_start_xmit(struct sk_buff *, struct net_device *);
+static int dscc4_close(struct net_device *);
+static int dscc4_ioctl(struct net_device *dev, struct ifreq *rq, int cmd);
+static int dscc4_change_mtu(struct net_device *dev, int mtu);
+static int dscc4_init_ring(struct net_device *);
+static void dscc4_release_ring(struct dscc4_dev_priv *);
+static void dscc4_timer(unsigned long);
+static void dscc4_tx_timeout(struct net_device *);
+static void dscc4_irq(int irq, void *dev_id, struct pt_regs *ptregs);
+static struct net_device_stats *dscc4_get_stats(struct net_device *);
+static int dscc4_attach_hdlc_device(struct net_device *);
+static void dscc4_unattach_hdlc_device(struct net_device *);
+static int dscc4_hdlc_open(struct hdlc_device_struct *);
+static void dscc4_hdlc_close(struct hdlc_device_struct *);
+static int dscc4_hdlc_ioctl(struct hdlc_device_struct *, struct ifreq *, int);
+static int dscc4_hdlc_xmit(hdlc_device *, struct sk_buff *);
+#ifdef EXPERIMENTAL_POLLING
+static int dscc4_tx_poll(struct dscc4_dev_priv *, struct net_device *);
+#endif
+
+void inline reset_TxFD(struct TxFD *tx_fd) {
+	/* FIXME: test with the last arg (size specification) = 0 */
+	tx_fd->state = FrameEnd | Hold | 0x00100000;
+	tx_fd->complete = 0x00000000;
+}
+
+void inline dscc4_release_ring_skbuff(struct sk_buff **p, int n)
+{
+	for(; n > 0; n--) {
+		if (*p)
+			dev_kfree_skb(*p);
+		p++;
+	}
+}
+
+static void dscc4_release_ring(struct dscc4_dev_priv *dpriv)
+{
+	struct pci_dev *pdev = dpriv->pci_priv->pdev;
+
+	pci_free_consistent(pdev, TX_RING_SIZE*sizeof(struct TxFD),
+			    dpriv->tx_fd, dpriv->tx_fd_dma);
+	pci_free_consistent(pdev, RX_RING_SIZE*sizeof(struct RxFD),
+			    dpriv->rx_fd, dpriv->rx_fd_dma);
+	dscc4_release_ring_skbuff(dpriv->tx_skbuff, TX_RING_SIZE);
+	dscc4_release_ring_skbuff(dpriv->rx_skbuff, RX_RING_SIZE);
+}
+
+void inline try_get_rx_skb(struct dscc4_dev_priv *priv, int cur, struct net_device *dev)
+{
+	struct sk_buff *skb;
+
+	skb = dev_alloc_skb(RX_MAX(dev->mtu+2));
+	priv->rx_skbuff[cur] = skb;
+	if (!skb) {
+		priv->rx_fd[cur--].data = (u32) NULL;
+		priv->rx_fd[cur%RX_RING_SIZE].state1 |= Hold;
+		priv->rx_needs_refill++;
+		return;
+	}
+	skb->dev = dev;
+	skb->protocol = htons(ETH_P_IP);
+	skb->mac.raw = skb->data;
+	priv->rx_fd[cur].data = pci_map_single(priv->pci_priv->pdev, skb->data,
+					       skb->len, PCI_DMA_FROMDEVICE);
+}
+
+/*
+ * IRQ/thread/whatever safe
+ */
+static int dscc4_wait_ack_cec(u32 ioaddr, struct net_device *dev, char *msg)
+{
+	s16 i = 0;
+
+	while (readl(ioaddr + STAR) & SccBusy) {
+		if (i++ < 0)  {
+			printk(KERN_ERR "%s: %s timeout\n", dev->name, msg);
+			return -1;
+		}
+	}
+	printk(KERN_DEBUG "%s: %s ack (%d try)\n", dev->name, msg, i);
+	return 0;
+}
+
+static int dscc4_do_action(struct net_device *dev, char *msg)
+{
+	unsigned long ioaddr = dev->base_addr;
+	u32 state;
+	s16 i;
+
+	writel(Action, ioaddr + GCMDR);
+	ioaddr += GSTAR;
+	for (i = 0; i >= 0; i++) {
+		state = readl(ioaddr);
+		if (state & Arf) {
+			printk(KERN_ERR "%s: %s failed\n", dev->name, msg);
+			writel(Arf, ioaddr);
+			return -1;
+		} else if (state & ArAck) {
+			printk(KERN_DEBUG "%s: %s ack (%d try)\n",
+			       dev->name, msg, i);
+			writel(ArAck, ioaddr);
+			return 0;
+		}
+	}
+	printk(KERN_ERR "%s: %s timeout\n", dev->name, msg);
+	return -1;
+}
+
+static __inline__ int dscc4_xpr_ack(struct dscc4_dev_priv *dpriv)
+{
+	int cur;
+	s16 i;
+
+	cur = dpriv->iqtx_current%IRQ_RING_SIZE;
+	for (i = 0; i >= 0; i++) {
+		if (!(dpriv->flags & (NeedIDR | NeedIDT)) ||
+		    (dpriv->iqtx[cur] & Xpr))
+			return 0;
+	}
+	printk(KERN_ERR "%s: %s timeout\n", "dscc4", "XPR");
+	return -1;
+}
+
+static __inline__ void dscc4_rx_skb(struct dscc4_dev_priv *dpriv, int cur,
+	struct RxFD *rx_fd, struct net_device *dev)
+{
+	struct pci_dev *pdev = dpriv->pci_priv->pdev;
+	struct sk_buff *skb;
+	int pkt_len;
+
+	skb = dpriv->rx_skbuff[cur];
+	pkt_len = TO_SIZE(rx_fd->state2);
+	pci_dma_sync_single(pdev, rx_fd->data, pkt_len, PCI_DMA_FROMDEVICE);
+	if((skb->data[pkt_len - 1] & FrameOk) == FrameOk) {
+		pci_unmap_single(pdev, rx_fd->data, skb->len, PCI_DMA_FROMDEVICE);
+		dpriv->stats.rx_packets++;
+		dpriv->stats.rx_bytes += pkt_len;
+		skb->tail += pkt_len;
+		skb->len = pkt_len;
+       		if (netif_running(hdlc_to_dev(&dpriv->hdlc)))
+			hdlc_netif_rx(&dpriv->hdlc, skb);
+		else
+			netif_rx(skb);
+		try_get_rx_skb(dpriv, cur, dev);
+	} else {
+		if(skb->data[pkt_len - 1] & FrameRdo)
+			dpriv->stats.rx_fifo_errors++;
+		else if(!(skb->data[pkt_len - 1] | ~FrameCrc))
+			dpriv->stats.rx_crc_errors++;
+		else if(!(skb->data[pkt_len - 1] | ~FrameVfr))
+			dpriv->stats.rx_length_errors++;
+		else
+			dpriv->stats.rx_errors++;
+	}
+	rx_fd->state1 |= Hold;
+	rx_fd->state2 = 0x00000000;
+	rx_fd->end = 0xbabeface;
+	if (!rx_fd->data)
+		return;
+	rx_fd--;
+	if (!cur)
+		rx_fd += RX_RING_SIZE;
+	rx_fd->state1 &= ~Hold;
+}
+
+static int __init dscc4_init_one (struct pci_dev *pdev,
+				  const struct pci_device_id *ent)
+{
+	struct dscc4_pci_priv *priv;
+	struct dscc4_dev_priv *dpriv;
+	int i;
+	static int cards_found = 0;
+	unsigned long ioaddr;
+
+	printk(KERN_DEBUG "%s", version);
+
+	if (pci_enable_device(pdev))
+		goto err_out;
+	if (!request_mem_region(pci_resource_start(pdev, 0),
+	                	pci_resource_len(pdev, 0), "registers")) {
+	        printk (KERN_ERR "dscc4: can't reserve MMIO region (regs)\n");
+	        goto err_out;
+	}
+	if (!request_mem_region(pci_resource_start(pdev, 1),
+	                        pci_resource_len(pdev, 1), "LBI interface")) {
+	        printk (KERN_ERR "dscc4: can't reserve MMIO region (lbi)\n");
+	        goto err_out_free_mmio_region0;
+	}
+	ioaddr = (unsigned long)ioremap(pci_resource_start(pdev, 0),
+					pci_resource_len(pdev, 0));
+	if (!ioaddr) {
+		printk(KERN_ERR "dscc4: cannot remap MMIO region %lx @ %lx\n",
+	                pci_resource_len(pdev, 0), pci_resource_start(pdev, 0));
+		goto err_out_free_mmio_region;
+	}
+	printk(KERN_DEBUG "Siemens DSCC4, MMIO at %#lx (regs), %#lx (lbi), IRQ %d.\n",
+	        pci_resource_start(pdev, 0),
+	        pci_resource_start(pdev, 1), pdev->irq);
+
+	/* High PCI latency useless. Cf app. note. */
+	pci_write_config_byte(pdev, PCI_LATENCY_TIMER, 0x10);
+	pci_set_master(pdev);
+
+	if (dscc4_found1(pdev, ioaddr))
+	        goto err_out_iounmap;
+
+	priv = (struct dscc4_pci_priv *)pci_get_drvdata(pdev);
+
+	if (request_irq(pdev->irq, &dscc4_irq, SA_SHIRQ, "dscc4", priv->root)) {
+		printk(KERN_WARNING "dscc4: IRQ %d is busy\n", pdev->irq);
+		goto err_out_iounmap;
+	}
+	priv->pdev = pdev;
+
+	/* power up/little endian/dma core controlled via hold bit */
+	writel(0x00000000, ioaddr + GMODE);
+	/* Shared interrupt queue */
+	{
+		u32 bits;
+
+		bits = (IRQ_RING_SIZE >> 5) - 1;
+		bits |= bits << 4;
+		bits |= bits << 8;
+		bits |= bits << 16;
+		writel(bits, ioaddr + IQLENR0);
+	}
+	/* Global interrupt queue */
+	writel((u32)(((IRQ_RING_SIZE >> 5) - 1) << 20), ioaddr + IQLENR1);
+	priv->iqcfg = (u32 *) pci_alloc_consistent(pdev,
+		IRQ_RING_SIZE*sizeof(u32), &priv->iqcfg_dma);
+	if (!priv->iqcfg)
+		goto err_out_free_irq;
+	writel(priv->iqcfg_dma, ioaddr + IQCFG);
+
+	/* 
+	 * SCC 0-3 private rx/tx irq structures 
+	 * IQRX/TXi needs to be set soon. Learned it the hard way...
+	 */
+	for(i = 0; i < dev_per_card; i++) {
+		dpriv = (struct dscc4_dev_priv *)(priv->root + i)->priv;
+		dpriv->iqtx = (u32 *) pci_alloc_consistent(pdev,
+			IRQ_RING_SIZE*sizeof(u32), &dpriv->iqtx_dma);
+		if (!dpriv->iqtx)
+			goto err_out_free_iqtx;
+		writel(dpriv->iqtx_dma, ioaddr + IQTX0 + i*4);
+	}
+	for(i = 0; i < dev_per_card; i++) {
+		dpriv = (struct dscc4_dev_priv *)(priv->root + i)->priv;
+		dpriv->iqrx = (u32 *) pci_alloc_consistent(pdev,
+			IRQ_RING_SIZE*sizeof(u32), &dpriv->iqrx_dma);
+		if (!dpriv->iqrx)
+			goto err_out_free_iqrx;
+		writel(dpriv->iqrx_dma, ioaddr + IQRX0 + i*4);
+	}
+
+	/* 
+	 * Cf application hint. Beware of hard-lock condition on 
+	 * threshold .
+	 */
+	writel(0x42104000, ioaddr + FIFOCR1);
+	//writel(0x9ce69800, ioaddr + FIFOCR2);
+	writel(0xdef6d800, ioaddr + FIFOCR2);
+	//writel(0x11111111, ioaddr + FIFOCR4);
+	writel(0x18181818, ioaddr + FIFOCR4);
+	// FIXME: should depend on the chipset revision
+	writel(0x0000000e, ioaddr + FIFOCR3);
+
+	writel(0xff200001, ioaddr + GCMDR);
+
+	cards_found++;
+	return 0;
+
+err_out_free_iqrx:
+	while (--i >= 0) {
+		dpriv = (struct dscc4_dev_priv *)(priv->root + i)->priv;
+		pci_free_consistent(pdev, IRQ_RING_SIZE*sizeof(u32), 
+				    dpriv->iqrx, dpriv->iqrx_dma);
+	}
+	i = dev_per_card;
+err_out_free_iqtx:
+	while (--i >= 0) {
+		dpriv = (struct dscc4_dev_priv *)(priv->root + i)->priv;
+		pci_free_consistent(pdev, IRQ_RING_SIZE*sizeof(u32), 
+				    dpriv->iqtx, dpriv->iqtx_dma);
+	}
+	pci_free_consistent(pdev, IRQ_RING_SIZE*sizeof(u32), priv->iqcfg, 
+			    priv->iqcfg_dma);
+err_out_free_irq:
+	free_irq(pdev->irq, priv->root);
+err_out_iounmap:
+	iounmap ((void *)ioaddr);
+err_out_free_mmio_region:
+	release_mem_region(pci_resource_start(pdev, 1),
+			   pci_resource_len(pdev, 1));
+err_out_free_mmio_region0:
+	release_mem_region(pci_resource_start(pdev, 0),
+			   pci_resource_len(pdev, 0));
+err_out:
+	return -ENODEV;
+};
+
+static int dscc4_found1(struct pci_dev *pdev, unsigned long ioaddr)
+{
+	struct dscc4_pci_priv *ppriv;
+	struct dscc4_dev_priv *dpriv;
+	struct net_device *dev;
+	int i = 0;
+
+	dpriv = (struct dscc4_dev_priv *)
+		kmalloc(dev_per_card*sizeof(struct dscc4_dev_priv), GFP_KERNEL);
+	if (!dpriv) {
+		printk(KERN_ERR "dscc4: can't allocate data\n");
+		goto err_out;
+	}
+	memset(dpriv, 0, dev_per_card*sizeof(struct dscc4_dev_priv));
+
+	dev = (struct net_device *)
+	      kmalloc(dev_per_card*sizeof(struct net_device), GFP_KERNEL);
+	if (!dev) {
+		printk(KERN_ERR "dscc4: can't allocate net_device\n");
+		goto err_dealloc_priv;
+	}
+	memset(dev, 0, dev_per_card*sizeof(struct net_device));
+
+	ppriv = (struct dscc4_pci_priv *)
+		kmalloc(sizeof(struct dscc4_pci_priv), GFP_KERNEL);
+	if (!ppriv) {
+		printk(KERN_ERR "dscc4: can't allocate pci private data.\n");
+		goto err_dealloc_dev;
+	}
+	memset(ppriv, 0, sizeof(struct dscc4_pci_priv));
+
+	for (i = 0; i < dev_per_card; i++) {
+		struct dscc4_dev_priv *p;
+		struct net_device *d;
+
+	        d = dev + i;
+	        d->base_addr = ioaddr;
+		d->init = NULL;
+	        d->irq = pdev->irq;
+		/* The card adds the crc */
+		d->type = ARPHRD_RAWHDLC;
+	        d->open = dscc4_open;
+	        d->stop = dscc4_close;
+	        d->hard_start_xmit = dscc4_start_xmit;
+		d->set_multicast_list = NULL;
+	        d->do_ioctl = dscc4_ioctl;
+		d->get_stats = dscc4_get_stats;
+		d->change_mtu = dscc4_change_mtu;
+	        d->mtu = HDLC_MAX_MTU;
+	        d->flags = IFF_MULTICAST|IFF_POINTOPOINT|IFF_NOARP;
+		d->tx_timeout = dscc4_tx_timeout;
+		d->watchdog_timeo = TX_TIMEOUT;
+
+		p = dpriv + i;
+		p->dev_id = i;
+		p->pci_priv = ppriv;
+		spin_lock_init(&p->lock);
+		d->priv = p;
+
+		if (dev_alloc_name(d, "scc%d")<0) {
+			printk(KERN_ERR "dev_alloc_name failed for scc.\n");
+			goto err_dealloc_dev;
+		}
+	        if (register_netdev(d)) {
+			printk(KERN_ERR "%s: register_netdev != 0.\n", d->name);
+			goto err_dealloc_dev;
+	        }
+		dscc4_attach_hdlc_device(d);
+		SET_MODULE_OWNER(d);
+	}
+	ppriv->root = dev;
+	ppriv->pdev = pdev;
+	spin_lock_init(&ppriv->lock);
+	pdev->driver_data = ppriv;
+	pci_set_drvdata(pdev, ppriv);
+	return 0;
+
+err_dealloc_dev:
+	while (--i >= 0)
+		unregister_netdev(dev + i);
+	kfree(dev);
+err_dealloc_priv:
+	kfree(dpriv);
+err_out:
+	return -1;
+};
+
+static void dscc4_timer(unsigned long data)
+{
+	struct net_device *dev = (struct net_device *)data;
+	struct dscc4_dev_priv *dpriv;
+	struct dscc4_pci_priv *ppriv;
+
+	dpriv = dev->priv;
+	if (netif_queue_stopped(dev) && 
+	   ((jiffies - dev->trans_start) > TX_TIMEOUT)) {
+		ppriv = dpriv->pci_priv;
+		if (dpriv->iqtx[dpriv->iqtx_current%IRQ_RING_SIZE]) {
+			u32 flags;
+
+			printk(KERN_DEBUG "%s: pending events\n", dev->name);
+			dev->trans_start = jiffies;
+			spin_lock_irqsave(&ppriv->lock, flags);
+			dscc4_tx_irq(ppriv, dev);
+			spin_unlock_irqrestore(&ppriv->lock, flags);
+		} else {
+			struct TxFD *tx_fd;
+			struct sk_buff *skb;
+			int i,j;
+
+			printk(KERN_DEBUG "%s: missing events\n", dev->name);
+			i = dpriv->tx_dirty%TX_RING_SIZE; 
+			j = dpriv->tx_current - dpriv->tx_dirty;
+			dpriv->stats.tx_dropped += j;
+			while(j--) {
+				skb = dpriv->tx_skbuff[i];
+				tx_fd = dpriv->tx_fd + i;
+				if (skb) {
+					dpriv->tx_skbuff[i] = NULL;
+					pci_unmap_single(ppriv->pdev, tx_fd->data, skb->len,
+						 	 PCI_DMA_TODEVICE);
+					dev_kfree_skb_irq(skb);
+				} else 
+					printk(KERN_INFO "%s: hardware on drugs!\n", dev->name);
+				tx_fd->data = 0; /* DEBUG */
+				tx_fd->complete &= ~DataComplete;
+				i++;	
+				i %= TX_RING_SIZE; 
+			}
+			dpriv->tx_dirty = dpriv->tx_current;
+			dev->trans_start = jiffies;
+			netif_wake_queue(dev);
+			printk(KERN_DEBUG "%s: re-enabled\n", dev->name);	
+		}
+	}
+        dpriv->timer.expires = jiffies + TX_TIMEOUT;
+        add_timer(&dpriv->timer);
+}
+
+static void dscc4_tx_timeout(struct net_device *dev)
+{
+	/* FIXME: something is missing there */
+};
+
+static int dscc4_open(struct net_device *dev)
+{
+	struct dscc4_dev_priv *dpriv = (struct dscc4_dev_priv *)dev->priv;
+	struct dscc4_pci_priv *ppriv;
+	u32 ioaddr = 0;
+
+	MOD_INC_USE_COUNT;
+
+	ppriv = dpriv->pci_priv;
+
+	if (dscc4_init_ring(dev))
+		goto err_out;
+
+	ioaddr = dev->base_addr + SCC_REG_START(dpriv->dev_id);
+
+	/* FIXME: VIS */
+	writel(readl(ioaddr + CCR0) | 0x80001000, ioaddr + CCR0);
+
+	writel(LengthCheck | (dev->mtu >> 5), ioaddr + RLCR);
+
+	/* no address recognition/crc-CCITT/cts enabled */
+	writel(readl(ioaddr + CCR1) | 0x021c8000, ioaddr + CCR1);
+
+	/* Ccr2.Rac = 0 */
+	writel(0x00050008 & ~RxActivate, ioaddr + CCR2);
+
+#ifdef EXPERIMENTAL_POLLING
+	writel(0xfffeef7f, ioaddr + IMR); /* Interrupt mask */
+#else
+	/* Don't mask RDO. Ever. */
+	//writel(0xfffaef7f, ioaddr + IMR); /* Interrupt mask */
+	writel(0xfffaef7e, ioaddr + IMR); /* Interrupt mask */
+#endif
+	/* IDT+IDR during XPR */
+	dpriv->flags = NeedIDR | NeedIDT;
+
+	/*
+	 * The following is a bit paranoid...
+	 *
+	 * NB: the datasheet "...CEC will stay active if the SCC is in
+	 * power-down mode or..." and CCR2.RAC = 1 are two different
+	 * situations.
+	 */
+	if (readl(ioaddr + STAR) & SccBusy) {
+		printk(KERN_ERR "%s busy. Try later\n", dev->name);
+		goto err_free_ring;
+	}
+	writel(TxSccRes | RxSccRes, ioaddr + CMDR);
+
+	/* ... the following isn't */
+	if (dscc4_wait_ack_cec(ioaddr, dev, "Cec"))
+		goto err_free_ring;
+
+	/* 
+	 * I would expect XPR near CE completion (before ? after ?).
+	 * At worst, this code won't see a late XPR and people
+	 * will have to re-issue an ifconfig (this is harmless). 
+	 * WARNING, a really missing XPR usually means a hardware 
+	 * reset is needed. Suggestions anyone ?
+	 */
+	if (dscc4_xpr_ack(dpriv))
+		goto err_free_ring;
+	
+	netif_start_queue(dev);
+
+        init_timer(&dpriv->timer);
+        dpriv->timer.expires = jiffies + 10*HZ;
+        dpriv->timer.data = (unsigned long)dev;
+        dpriv->timer.function = &dscc4_timer;
+        add_timer(&dpriv->timer);
+	netif_carrier_on(dev);
+
+	return 0;
+
+err_free_ring:
+	dscc4_release_ring(dpriv);
+err_out:
+	MOD_DEC_USE_COUNT;
+	return -EAGAIN;
+}
+
+#ifdef EXPERIMENTAL_POLLING
+static int dscc4_tx_poll(struct dscc4_dev_priv *dpriv, struct net_device *dev)
+{
+	/* FIXME: it's gonna be easy (TM), for sure */
+}
+#endif /* EXPERIMENTAL_POLLING */
+
+static int dscc4_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	struct dscc4_dev_priv *dpriv = dev->priv;
+	struct dscc4_pci_priv *ppriv;
+	struct TxFD *tx_fd;
+	int cur, next;
+
+	ppriv = dpriv->pci_priv;
+	cur = dpriv->tx_current++%TX_RING_SIZE;
+	next = dpriv->tx_current%TX_RING_SIZE;
+	dpriv->tx_skbuff[next] = skb;
+	tx_fd = dpriv->tx_fd + next;
+	tx_fd->state = FrameEnd | Hold | TO_STATE(skb->len & TxSizeMax);
+	tx_fd->data = pci_map_single(ppriv->pdev, skb->data, skb->len,
+				     PCI_DMA_TODEVICE);
+	tx_fd->complete = 0x00000000;
+	mb(); // FIXME: suppress ?
+
+#ifdef EXPERIMENTAL_POLLING
+	spin_lock(&dpriv->lock);
+	while(dscc4_tx_poll(dpriv, dev));
+	spin_unlock(&dpriv->lock);
+#endif
+	/*
+	 * I know there's a window for a race in the following lines but
+	 * dscc4_timer will take good care of it. The chipset eats events
+	 * (especially the net_dev re-enabling ones) thus there is no
+	 * reason to try and be smart.
+	 */
+	if ((dpriv->tx_dirty + 16) < dpriv->tx_current) {
+			netif_stop_queue(dev);
+			dpriv->hi_expected = 2;
+	}
+	tx_fd = dpriv->tx_fd + cur;
+	tx_fd->state &= ~Hold;
+	mb(); // FIXME: suppress ?
+
+	/* 
+	 * One may avoid some pci transactions during intense TX periods.
+	 * Not sure it's worth the pain...
+	 */
+	writel((TxPollCmd << dpriv->dev_id) | NoAck, dev->base_addr + GCMDR);
+	dev->trans_start = jiffies;
+	return 0;
+}
+
+static int dscc4_close(struct net_device *dev)
+{
+	struct dscc4_dev_priv *dpriv = (struct dscc4_dev_priv *)dev->priv;
+	u32 ioaddr = dev->base_addr;
+	int dev_id;
+
+	del_timer_sync(&dpriv->timer);
+	netif_stop_queue(dev);
+
+	dev_id = dpriv->dev_id;
+
+	writel(0x00050000, ioaddr + SCC_REG_START(dev_id) + CCR2);
+	writel(MTFi|Rdr|Rdt, ioaddr + CH0CFG + dev_id*0x0c); /* Reset Rx/Tx */
+	writel(0x00000001, ioaddr + GCMDR);
+
+	dscc4_release_ring(dpriv);
+
+	MOD_DEC_USE_COUNT;
+	return 0;
+}
+
+static int dscc4_set_clock(struct net_device *dev, u32 *bps, u32 *state)
+{
+	struct dscc4_dev_priv *dpriv = (struct dscc4_dev_priv *)dev->priv;
+	u32 brr;
+
+	*state &= ~Ccr0ClockMask;
+	if (*bps) { /* DCE */
+		u32 n = 0, m = 0, divider;
+		int xtal;
+
+		xtal = dpriv->pci_priv->xtal_hz;
+		if (!xtal)
+			return -1;
+		divider = xtal / *bps;
+		if (divider > BRR_DIVIDER_MAX) {
+			divider >>= 4;
+			*state |= 0x00000036; /* Clock mode 6b (BRG/16) */
+		} else
+			*state |= 0x00000037; /* Clock mode 7b (BRG) */
+		if (divider >> 22) {
+			n = 63;
+			m = 15;
+		} else if (divider) {
+			/* Extraction of the 6 highest weighted bits */
+			m = 0;
+			while (0xffffffc0 & divider) {
+				m++;
+				divider >>= 1;
+			}
+			n = divider;
+		}
+		brr = (m << 8) | n;
+		divider = n << m;
+		if (!(*state & 0x00000001)) /* Clock mode 6b */
+			divider <<= 4;
+		*bps = xtal / divider;
+	} else { /* DTE */
+		/* 
+		 * "state" already reflects Clock mode 0a. 
+		 * Nothing more to be done 
+		 */
+		brr = 0;
+	}
+	writel(brr, dev->base_addr + BRR + SCC_REG_START(dpriv->dev_id));
+
+	return 0;
+}
+
+#ifdef LATER_PLEASE
+/*
+ * -*- [RFC] Configuring Synchronous Interfaces in Linux -*-
+ */
+
+// FIXME: MEDIA already defined in linux/hdlc.h
+#define HDLC_MEDIA_V35		0
+#define HDLC_MEDIA_RS232	1
+#define HDLC_MEDIA_X21		2
+#define HDLC_MEDIA_E1		3
+#define HDLC_MEDIA_HSSI		4
+
+#define HDLC_CODING_NRZ		0
+#define HDLC_CODING_NRZI	1
+#define HDLC_CODING_FM0		2
+#define HDLC_CODING_FM1		3
+#define HDLC_CODING_MANCHESTER	4
+
+#define HDLC_CRC_NONE		0
+#define HDLC_CRC_16		1
+#define HDLC_CRC_32		2
+#define HDLC_CRC_CCITT		3
+
+/* RFC: add the crc reset value ? */
+struct hdlc_physical {
+	u8 media;
+	u8 coding;
+	u32 rate;
+	u8 crc;
+	u8 crc_siz;		/* 2 or 4 bytes */
+	u8 shared_flags;	/* Discouraged on the DSCC4 */
+};
+
+// FIXME: PROTO already defined in linux/hdlc.h
+#define HDLC_PROTO_RAW		0
+#define HDLC_PROTO_FR		1
+#define HDLC_PROTO_X25		2
+#define HDLC_PROTO_PPP		3
+#define HDLC_PROTO_CHDLC	4
+
+struct hdlc_protocol {
+	u8 proto; 
+	
+	union {
+	} u;
+};
+
+struct screq {
+	u16 media_group;
+
+	union {
+		struct hdlc_physical hdlc_phy;
+		struct hdlc_protocol hdlc_proto;
+	} u;
+};
+
+// FIXME: go sub-module 
+static struct {
+	u16 coding;
+	u16 bits;
+} map[] = {
+	{HDLC_CODING_NRZ,		0x00},
+	{HDLC_CODING_NRZI,		0x20},
+	{HDLC_CODING_FM0,		0x40},
+	{HDLC_CODING_FM1,		0x50},
+	{HDLC_CODING_MANCHESTER,	0x60},
+	{65535,	0x00}
+};
+#endif /* LATER_PLEASE */
+
+static int dscc4_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
+{
+	struct dscc4_dev_priv *dpriv = dev->priv;
+	u32 state, ioaddr;
+
+        if (dev->flags & IFF_UP)
+                return -EBUSY;
+
+	switch (cmd) {
+		/* Set built-in quartz frequency */
+		case SIOCDEVPRIVATE: {
+			u32 hz;
+
+			hz = ifr->ifr_ifru.ifru_ivalue;
+			if (hz >= 33000000) /* 33 MHz */
+				return -EOPNOTSUPP;
+			dpriv->pci_priv->xtal_hz = hz;
+			return 0;
+		}
+		/* Set/unset loopback */
+		case SIOCDEVPRIVATE+1: {
+			u32 flags;
+	
+			ioaddr = dev->base_addr + CCR1 +
+				 SCC_REG_START(dpriv->dev_id);
+			state = readl(ioaddr);
+			flags = ifr->ifr_ifru.ifru_ivalue;
+			if (flags & 0x00000001) {
+				printk(KERN_DEBUG "%s: loopback\n", dev->name);
+				state |= 0x00000100;
+			} else {
+				printk(KERN_DEBUG "%s: normal\n", dev->name);
+				state &= ~0x00000100;
+			}
+			writel(state, ioaddr);
+			return 0;
+		}
+
+#ifdef LATER_PLEASE
+		case SIOCDEVPRIVATE+2: {
+		{
+			struct screq scr;
+
+			err = copy_from_user(&scr, ifr->ifr_ifru.ifru_data, sizeof(struct screq));
+			if (err)
+				return err;
+			do {
+				if (scr.u.hdlc_phy.coding == map[i].coding)
+					break;
+			} while (map[++i].coding != 65535);
+			if (!map[i].coding)
+				return -EOPNOTSUPP;
+			
+			ioaddr = dev->base_addr + CCR0 +
+				 SCC_REG_START(dpriv->dev_id);
+			state = readl(ioaddr) & ~EncodingMask;
+			state |= (u32)map[i].bits << 16;
+			writel(state, ioaddr);	
+			printk("state: %08x\n", state); /* DEBUG */
+			return 0;
+		}
+		case SIOCDEVPRIVATE+3: {
+			struct screq *scr = (struct screq *)ifr->ifr_ifru.ifru_data;
+
+			ioaddr = dev->base_addr + CCR0 +
+				 SCC_REG_START(dpriv->dev_id);
+			state = (readl(ioaddr) & EncodingMask) >> 16;
+			do {
+				if (state == map[i].bits)
+					break;
+			} while (map[++i].coding);
+			return put_user(map[i].coding, (u16 *)scr->u.hdlc_phy.coding);
+		}
+#endif /* LATER_PLEASE */
+
+		case HDLCSCLOCKRATE:
+		{
+			u32 state, bps;
+ 
+			bps = ifr->ifr_ifru.ifru_ivalue;
+			ioaddr = dev->base_addr + CCR0 +
+				 SCC_REG_START(dpriv->dev_id);
+			state = readl(ioaddr);
+			if(dscc4_set_clock(dev, &bps, &state) < 0)
+				return -EOPNOTSUPP;
+			if (bps) { /* DCE */
+				printk(KERN_DEBUG "%s: generated RxClk (DCE)\n",
+				       dev->name);
+				ifr->ifr_ifru.ifru_ivalue = bps;
+			} else { /* DTE */
+				state = 0x80001000;
+				printk(KERN_DEBUG "%s: external RxClk (DTE)\n",
+				       dev->name);
+			}
+			writel(state, ioaddr);	
+			return 0;
+		}
+		case HDLCGCLOCKRATE: {
+			u32 brr;
+			int bps;
+
+			brr = readl(dev->base_addr + BRR + 
+				    SCC_REG_START(dpriv->dev_id));
+			bps = dpriv->pci_priv->xtal_hz >> (brr >> 8);
+			bps /= (brr & 0x3f) + 1;
+			ifr->ifr_ifru.ifru_ivalue = bps;
+			return 0;
+		}
+
+		default:
+			return -EOPNOTSUPP;
+	}
+}
+
+static int dscc4_change_mtu(struct net_device *dev, int mtu)
+{
+	/* FIXME: chainsaw coded... */
+	if ((mtu <= 3) || (mtu > 65531))
+		return -EINVAL;
+        if(dev->flags & IFF_UP)
+                return -EBUSY;
+	dev->mtu = mtu;
+        return(0);
+}
+
+static void dscc4_irq(int irq, void *dev_instance, struct pt_regs *ptregs)
+{
+	struct net_device *dev = dev_instance;
+	struct dscc4_pci_priv *priv;
+	u32 ioaddr, state;
+	unsigned long flags;
+	int i;
+
+	priv = ((struct dscc4_dev_priv *)dev->priv)->pci_priv;
+	/* 
+	 * FIXME: shorten the protected area (set some bit telling we're
+	 * in an interrupt or increment some work-to-do counter etc...)
+	 */
+	spin_lock_irqsave(&priv->lock, flags);
+
+	ioaddr = dev->base_addr;
+
+	state = readl(ioaddr + GSTAR);
+	if (!state)
+		goto out;
+	writel(state, ioaddr + GSTAR);
+
+	if (state & Arf) { 
+		printk(KERN_ERR "%s: failure (Arf). Harass the maintener\n",
+		       dev->name);
+		goto out;
+	}
+	state &= ~ArAck;
+	if (state & Cfg) {
+		if (debug)
+			printk(KERN_DEBUG "CfgIV\n");
+		if (priv->iqcfg[priv->cfg_cur++%IRQ_RING_SIZE] & Arf)
+			printk(KERN_ERR "%s: %s failed\n", dev->name, "CFG");
+		if (!(state &= ~Cfg))
+			goto out;
+	}
+	if (state & RxEvt) {
+		i = dev_per_card - 1;
+		do {
+			dscc4_rx_irq(priv, dev + i);
+		} while (--i >= 0);
+		state &= ~RxEvt;
+	}
+	if (state & TxEvt) {
+		i = dev_per_card - 1;
+		do {
+			dscc4_tx_irq(priv, dev + i);
+		} while (--i >= 0);
+		state &= ~TxEvt;
+	}
+out:
+	spin_unlock_irqrestore(&priv->lock, flags);
+}
+
+static __inline__ void dscc4_tx_irq(struct dscc4_pci_priv *ppriv, 
+				    struct net_device *dev)
+{
+	struct dscc4_dev_priv *dpriv = dev->priv;
+	u32 state;
+	int cur, loop = 0;
+
+try:
+	cur = dpriv->iqtx_current%IRQ_RING_SIZE;
+	state = dpriv->iqtx[cur];
+	if (!state) {
+#ifdef DEBUG
+		if (loop > 1)
+			printk(KERN_DEBUG "%s: Tx irq loop=%d\n", dev->name, loop);
+#endif
+		if (loop && netif_queue_stopped(dev))
+			if ((dpriv->tx_dirty + 8) >= dpriv->tx_current)
+				netif_wake_queue(dev);
+		return;
+	}
+	loop++;
+	dpriv->iqtx[cur] = 0;
+	dpriv->iqtx_current++;
+
+#ifdef DEBUG_PARANOID
+	if (SOURCE_ID(state) != dpriv->dev_id) {
+		printk(KERN_DEBUG "%s (Tx): Source Id=%d, state=%08x\n",
+		       dev->name, SOURCE_ID(state), state );
+		return;
+	}
+	if (state & 0x0df80c01) {
+		printk(KERN_DEBUG "%s (Tx): state=%08x (UFO alert)\n",
+		       dev->name, state);
+		return;
+	}
+#endif
+	// state &= 0x0fffffff; /* Tracking the analyzed bits */
+	if (state & SccEvt) {
+		if (state & Alls) {
+			struct TxFD *tx_fd;
+			struct sk_buff *skb;
+
+			cur = dpriv->tx_dirty%TX_RING_SIZE;
+			tx_fd = dpriv->tx_fd + cur;
+
+			skb = dpriv->tx_skbuff[cur];
+
+			/* XXX: hideous kludge - to be removed "later" */
+			if (!skb) {
+				printk(KERN_ERR "%s: NULL skb in tx_irq at index %d\n", dev->name, cur);
+				goto try;
+			}
+			dpriv->tx_dirty++; // MUST be after skb test
+
+			/* Happens sometime. Don't know what triggers it */
+			if (!(tx_fd->complete & DataComplete)) {
+				u32 ioaddr, isr;
+
+				ioaddr = dev->base_addr + 
+					 SCC_REG_START(dpriv->dev_id) + ISR;
+				isr = readl(ioaddr);
+				printk(KERN_DEBUG 
+				       "%s: DataComplete=0 cur=%d isr=%08x state=%08x\n",
+				       dev->name, cur, isr, state);
+				writel(isr, ioaddr);
+				dpriv->stats.tx_dropped++;
+			} else {
+				tx_fd->complete &= ~DataComplete;
+				if (tx_fd->state & FrameEnd) {
+					dpriv->stats.tx_packets++;
+					dpriv->stats.tx_bytes += skb->len;
+				}
+			}
+
+			dpriv->tx_skbuff[cur] = NULL;
+			pci_unmap_single(ppriv->pdev, tx_fd->data, skb->len,
+					 PCI_DMA_TODEVICE);
+			tx_fd->data = 0; /* DEBUG */
+			dev_kfree_skb_irq(skb);
+{ // DEBUG
+			cur = (dpriv->tx_dirty-1)%TX_RING_SIZE;
+			tx_fd = dpriv->tx_fd + cur;
+			tx_fd->state |= Hold;
+}
+			if (!(state &= ~Alls))
+				goto try;
+		}
+		/* 
+		 * Transmit Data Underrun
+		 */
+		if (state & Xdu) {
+			printk(KERN_ERR "dscc4: XDU. Contact maintainer\n");
+			dpriv->flags = NeedIDT; 
+			/* Tx reset */
+			writel(MTFi | Rdt, 
+			       dev->base_addr + 0x0c*dpriv->dev_id + CH0CFG);
+			writel(0x00000001, dev->base_addr + GCMDR);
+			return;
+		}
+		if (state & Xmr) {
+			/* Frame needs to be sent again - FIXME */
+			//dscc4_start_xmit(dpriv->tx_skbuff[dpriv->tx_dirty], dev);
+			if (!(state &= ~0x00002000)) /* DEBUG */
+				goto try;
+		}
+		if (state & Xpr) {
+			unsigned long ioaddr = dev->base_addr;
+			unsigned long scc_offset;
+			u32 scc_addr;
+
+			scc_offset = ioaddr + SCC_REG_START(dpriv->dev_id);
+			scc_addr = ioaddr + 0x0c*dpriv->dev_id;
+			if (readl(scc_offset + STAR) & SccBusy)
+				printk(KERN_DEBUG "%s busy. Fatal\n", 
+				       dev->name);
+			/*
+			 * Keep this order: IDT before IDR
+			 */
+			if (dpriv->flags & NeedIDT) {
+				writel(MTFi | Idt, scc_addr + CH0CFG);
+				writel(dpriv->tx_fd_dma + 
+				       (dpriv->tx_dirty%TX_RING_SIZE)*
+				       sizeof(struct TxFD), scc_addr + CH0BTDA);
+				if(dscc4_do_action(dev, "IDT"))
+					goto err_xpr;
+				dpriv->flags &= ~NeedIDT;
+				mb();
+			}
+			if (dpriv->flags & NeedIDR) {
+				writel(MTFi | Idr, scc_addr + CH0CFG);
+				writel(dpriv->rx_fd_dma + 
+				       (dpriv->rx_current%RX_RING_SIZE)*
+				       sizeof(struct RxFD), scc_addr + CH0BRDA);
+				if(dscc4_do_action(dev, "IDR"))
+					goto err_xpr;
+				dpriv->flags &= ~NeedIDR;
+				mb();
+				/* Activate receiver and misc */
+				writel(0x08050008, scc_offset + CCR2);
+			}
+		err_xpr:
+			if (!(state &= ~Xpr))
+				goto try;
+		}
+	} else { /* ! SccEvt */
+		if (state & Hi) {
+#ifdef EXPERIMENTAL_POLLING
+			while(!dscc4_tx_poll(dpriv, dev));
+#endif
+			state &= ~Hi;
+		}
+		/*
+		 * FIXME: it may be avoided. Re-re-re-read the manual.
+		 */
+		if (state & Err) {
+			printk(KERN_ERR "%s: Tx ERR\n", dev->name);
+			dpriv->stats.tx_errors++;
+			state &= ~Err;
+		}
+	}
+	goto try;
+}
+
+static __inline__ void dscc4_rx_irq(struct dscc4_pci_priv *priv, struct net_device *dev)
+{
+	struct dscc4_dev_priv *dpriv = dev->priv;
+	u32 state;
+	int cur;
+
+try:
+	cur = dpriv->iqrx_current%IRQ_RING_SIZE;
+	state = dpriv->iqrx[cur];
+	if (!state)
+		return;
+	dpriv->iqrx[cur] = 0;
+	dpriv->iqrx_current++;
+
+#ifdef DEBUG_PARANOID
+	if (SOURCE_ID(state) != dpriv->dev_id) {
+		printk(KERN_DEBUG "%s (Rx): Source Id=%d, state=%08x\n",
+		       dev->name, SOURCE_ID(state), state);
+		goto try;
+	}
+	if (state & 0x0df80c01) {
+		printk(KERN_DEBUG "%s (Rx): state=%08x (UFO alert)\n",
+		       dev->name, state);
+		goto try;
+	}
+#endif
+	if (!(state & SccEvt)){
+		struct RxFD *rx_fd;
+
+		state &= 0x00ffffff;
+		if (state & Err) { /* Hold or reset */
+			printk(KERN_DEBUG "%s (Rx): ERR\n", dev->name);
+			cur = dpriv->rx_current;
+			rx_fd = dpriv->rx_fd + cur;
+			/*
+			 * Presume we're not facing a DMAC receiver reset. 
+			 * As We use the rx size-filtering feature of the 
+			 * DSCC4, the beginning of a new frame is waiting in 
+			 * the rx fifo. I bet a Receive Data Overflow will 
+			 * happen most of time but let's try and avoid it.
+			 * Btw (as for RDO) if one experiences ERR whereas
+			 * the system looks rather idle, there may be a 
+			 * problem with latency. In this case, increasing
+			 * RX_RING_SIZE may help.
+			 */
+			while (dpriv->rx_needs_refill) {
+				while(!(rx_fd->state1 & Hold)) {
+					rx_fd++;
+					cur++;
+					if (!(cur = cur%RX_RING_SIZE))
+						rx_fd = dpriv->rx_fd;
+				}
+				dpriv->rx_needs_refill--;
+				try_get_rx_skb(dpriv, cur, dev);
+				if (!rx_fd->data)
+					goto try;
+				rx_fd->state1 &= ~Hold;
+				rx_fd->state2 = 0x00000000;
+				rx_fd->end = 0xbabeface;
+			}
+			goto try;
+		}
+		if (state & Fi) {
+			cur = dpriv->rx_current%RX_RING_SIZE;
+			rx_fd = dpriv->rx_fd + cur;
+			dscc4_rx_skb(dpriv, cur, rx_fd, dev);
+			dpriv->rx_current++;
+			goto try;
+		}
+		if (state & Hi ) { /* HI bit */
+			state &= ~Hi;
+			goto try;
+		}
+	} else { /* ! SccEvt */
+#ifdef DEBUG_PARANOIA
+		int i;
+		static struct {
+			u32 mask;
+			const char *irq_name;
+		} evts[] = {
+			{ 0x00008000, "TIN"},
+			{ 0x00004000, "CSC"},
+			{ 0x00000020, "RSC"},
+			{ 0x00000010, "PCE"},
+			{ 0x00000008, "PLLA"},
+			{ 0x00000004, "CDSC"},
+			{ 0, NULL}
+		};
+#endif /* DEBUG_PARANOIA */
+		state &= 0x00ffffff;
+#ifdef DEBUG_PARANOIA
+		for (i = 0; evts[i].irq_name; i++) {
+			if (state & evts[i].mask) {
+				printk(KERN_DEBUG "dscc4(%s): %s\n",
+					dev->name, evts[i].irq_name);
+				if (!(state &= ~evts[i].mask))
+					goto try;
+			}
+		}
+#endif /* DEBUG_PARANOIA */
+		/*
+		 * Receive Data Overflow (FIXME: untested)
+		 */
+		if (state & Rdo) {
+			u32 ioaddr, scc_offset, scc_addr;
+			struct RxFD *rx_fd;
+			int cur;
+
+			//if (debug)
+			//	dscc4_rx_dump(dpriv);
+			ioaddr = dev->base_addr;
+			scc_addr = ioaddr + 0x0c*dpriv->dev_id;
+			scc_offset = ioaddr + SCC_REG_START(dpriv->dev_id);
+
+			writel(readl(scc_offset + CCR2) & ~RxActivate, 
+			       scc_offset + CCR2);
+			/*
+			 * This has no effect. Why ?
+			 * ORed with TxSccRes, one sees the CFG ack (for
+			 * the TX part only).
+			 */	
+			writel(RxSccRes, scc_offset + CMDR);
+			dpriv->flags |= RdoSet;
+
+			/* 
+			 * Let's try and save something in the received data.
+			 * rx_current must be incremented at least once to
+			 * avoid HOLD in the BRDA-to-be-pointed desc.
+			 */
+			do {
+				cur = dpriv->rx_current++%RX_RING_SIZE;
+				rx_fd = dpriv->rx_fd + cur;
+				if (!(rx_fd->state2 & DataComplete))
+					break;
+				if (rx_fd->state2 & FrameAborted) {
+					dpriv->stats.rx_over_errors++;
+					rx_fd->state1 |= Hold;
+					rx_fd->state2 = 0x00000000;
+					rx_fd->end = 0xbabeface;
+				} else 
+					dscc4_rx_skb(dpriv, cur, rx_fd, dev);
+			} while (1);
+
+			if (debug) {
+				if (dpriv->flags & RdoSet)
+					printk(KERN_DEBUG 
+					       "dscc4: no RDO in Rx data\n");
+			}
+#ifdef DSCC4_RDO_EXPERIMENTAL_RECOVERY
+			/*
+			 * FIXME: must the reset be this violent ?
+			 */
+			writel(dpriv->rx_fd_dma + 
+			       (dpriv->rx_current%RX_RING_SIZE)*
+			       sizeof(struct RxFD), scc_addr + CH0BRDA);
+			writel(MTFi|Rdr|Idr, scc_addr + CH0CFG);
+			if(dscc4_do_action(dev, "RDR")) {
+				printk(KERN_ERR "%s: RDO recovery failed(%s)\n",
+				       dev->name, "RDR");
+				goto rdo_end;
+			}
+			writel(MTFi|Idr, scc_addr + CH0CFG);
+			if(dscc4_do_action(dev, "IDR")) {
+				printk(KERN_ERR "%s: RDO recovery failed(%s)\n",
+				       dev->name, "IDR");
+				goto rdo_end;
+			}
+		rdo_end:
+#endif
+			writel(readl(scc_offset + CCR2) | RxActivate, 
+			       scc_offset + CCR2);
+			goto try;
+		}
+		/* These will be used later */
+		if (state & Rfs) {
+			if (!(state &= ~Rfs))
+				goto try;
+		}
+		if (state & Rfo) {
+			if (!(state &= ~Rfo))
+				goto try;
+		}
+		if (state & Flex) {
+			if (!(state &= ~Flex))
+				goto try;
+		}
+	}
+}
+
+static int dscc4_init_ring(struct net_device *dev)
+{
+	struct dscc4_dev_priv *dpriv = (struct dscc4_dev_priv *)dev->priv;
+	struct TxFD *tx_fd;
+	struct RxFD *rx_fd;
+	int i;
+
+	tx_fd = (struct TxFD *) pci_alloc_consistent(dpriv->pci_priv->pdev,
+		TX_RING_SIZE*sizeof(struct TxFD), &dpriv->tx_fd_dma);
+	if (!tx_fd)
+		goto err_out;
+	rx_fd = (struct RxFD *) pci_alloc_consistent(dpriv->pci_priv->pdev,
+		RX_RING_SIZE*sizeof(struct RxFD), &dpriv->rx_fd_dma);
+	if (!rx_fd)
+		goto err_free_dma_tx;
+
+	dpriv->tx_fd = tx_fd;
+	dpriv->rx_fd = rx_fd;
+	dpriv->rx_current = 0;
+	dpriv->tx_current = 0;
+	dpriv->tx_dirty = 0;
+
+	/* the dma core of the dscc4 will be locked on the first desc */
+	for(i = 0; i < TX_RING_SIZE; ) {
+		reset_TxFD(tx_fd);
+	        /* FIXME: NULL should be ok - to be tried */
+	        tx_fd->data = dpriv->tx_fd_dma;
+	        dpriv->tx_skbuff[i] = NULL;
+		i++;
+		tx_fd->next = (u32)(dpriv->tx_fd_dma + i*sizeof(struct TxFD));
+		tx_fd++;
+	}
+	(--tx_fd)->next = (u32)dpriv->tx_fd_dma;
+{
+	/*
+	 * XXX: I would expect the following to work for the first descriptor
+	 * (tx_fd->state = 0xc0000000)
+	 * - Hold=1 (don't try and branch to the next descripto);
+	 * - No=0 (I want an empty data section, i.e. size=0);
+	 * - Fe=1 (required by No=0 or we got an Err irq and must reset).
+	 * Alas, it fails (and locks solid). Thus the introduction of a dummy
+	 * skb to avoid No=0 (choose one: Ugly [ ] Tasteless [ ] VMS [ ]).
+	 * TODO: fiddle the tx threshold when time permits.
+	 */
+	struct sk_buff *skb;
+
+	skb = dev_alloc_skb(32);
+	if (!skb)
+		goto err_free_dma_tx;
+	skb->len = 32;
+	memset(skb->data, 0xaa, 16);
+	tx_fd -= (TX_RING_SIZE - 1);
+	tx_fd->state = 0xc0000000;
+	tx_fd->state |= ((u32)(skb->len & TxSizeMax)) << 16;
+	tx_fd->data = pci_map_single(dpriv->pci_priv->pdev, skb->data, 
+				     skb->len, PCI_DMA_TODEVICE);
+	dpriv->tx_skbuff[0] = skb;
+}
+	for (i = 0; i < RX_RING_SIZE;) {
+		/* size set by the host. Multiple of 4 bytes please */
+	        rx_fd->state1 = HiDesc; /* Hi, no Hold */
+	        rx_fd->state2 = 0x00000000;
+	        rx_fd->end = 0xbabeface;
+	        rx_fd->state1 |= ((u32)(dev->mtu & RxSizeMax)) << 16;
+		try_get_rx_skb(dpriv, i, dev);
+		i++;
+		rx_fd->next = (u32)(dpriv->rx_fd_dma + i*sizeof(struct RxFD));
+		rx_fd++;
+	}
+	(--rx_fd)->next = (u32)dpriv->rx_fd_dma;
+	rx_fd->state1 |= 0x40000000; /* Hold */
+
+	return 0;
+
+err_free_dma_tx:
+	pci_free_consistent(dpriv->pci_priv->pdev, TX_RING_SIZE*sizeof(*tx_fd),
+			    tx_fd, dpriv->tx_fd_dma);
+err_out:
+	return -1;
+}
+
+static struct net_device_stats *dscc4_get_stats(struct net_device *dev)
+{
+	struct dscc4_dev_priv *priv = (struct dscc4_dev_priv *)dev->priv;
+
+	return &priv->stats;
+}
+
+static void __exit dscc4_remove_one(struct pci_dev *pdev)
+{
+	struct dscc4_pci_priv *ppriv;
+	struct net_device *root;
+	int i;
+
+	ppriv = pci_get_drvdata(pdev);
+	root = ppriv->root;
+
+	free_irq(pdev->irq, root);
+	pci_free_consistent(pdev, IRQ_RING_SIZE*sizeof(u32), ppriv->iqcfg, 
+			    ppriv->iqcfg_dma);
+	for (i=0; i < dev_per_card; i++) {
+		struct dscc4_dev_priv *dpriv;
+		struct net_device *dev;
+
+		dev = ppriv->root + i;
+		dscc4_unattach_hdlc_device(dev);
+
+		dpriv = (struct dscc4_dev_priv *)dev->priv;
+		pci_free_consistent(pdev, IRQ_RING_SIZE*sizeof(u32), 
+				    dpriv->iqrx, dpriv->iqrx_dma);
+		pci_free_consistent(pdev, IRQ_RING_SIZE*sizeof(u32), 
+				    dpriv->iqtx, dpriv->iqtx_dma);
+		unregister_netdev(dev);
+	}
+	kfree(root->priv);
+
+	iounmap((void *)root->base_addr);
+	kfree(root);
+
+	kfree(ppriv);
+
+	release_mem_region(pci_resource_start(pdev, 1),
+			   pci_resource_len(pdev, 1));
+	release_mem_region(pci_resource_start(pdev, 0),
+			   pci_resource_len(pdev, 0));
+}
+
+static int dscc4_hdlc_ioctl(struct hdlc_device_struct *hdlc, struct ifreq *ifr, int cmd)
+{
+	struct net_device *dev = (struct net_device *)hdlc->netdev.base_addr;
+	int result;
+
+	/* FIXME: locking ? */
+	result = dscc4_ioctl(dev, ifr, cmd);
+	return result;
+}
+
+static int dscc4_hdlc_open(struct hdlc_device_struct *hdlc)
+{
+	struct net_device *dev = (struct net_device *)(hdlc->netdev.base_addr);
+
+        if (netif_running(dev)) {
+		printk(KERN_DEBUG "%s: already running\n", dev->name); // DEBUG
+		return 0;
+	}
+	return dscc4_open(dev);
+}
+
+static int dscc4_hdlc_xmit(hdlc_device *hdlc, struct sk_buff *skb)
+{
+	struct net_device *dev = (struct net_device *)hdlc->netdev.base_addr;
+
+	return dscc4_start_xmit(skb, dev);
+}
+
+static void dscc4_hdlc_close(struct hdlc_device_struct *hdlc)
+{
+	struct net_device *dev = (struct net_device *)hdlc->netdev.base_addr;
+	struct dscc4_dev_priv *dpriv;
+
+	dpriv = dev->priv;
+	--dpriv->usecount;
+}
+
+/* Operated under dev lock */
+static int dscc4_attach_hdlc_device(struct net_device *dev) 
+{
+	struct dscc4_dev_priv *dpriv = dev->priv;
+	struct hdlc_device_struct *hdlc;
+	int result;
+
+	hdlc = &dpriv->hdlc;
+	/* XXX: Don't look at the next line */
+	hdlc->netdev.base_addr = (unsigned long)dev;
+	// FIXME: set hdlc->set_mode ?
+	hdlc->open = dscc4_hdlc_open;
+	hdlc->close = dscc4_hdlc_close;
+	hdlc->ioctl = dscc4_hdlc_ioctl;
+	hdlc->xmit = dscc4_hdlc_xmit;
+
+	result = register_hdlc_device(hdlc);
+	if (!result)
+		dpriv->usecount++;
+	return result;
+}
+
+/* Operated under dev lock */
+static void dscc4_unattach_hdlc_device(struct net_device *dev) 
+{
+	struct dscc4_dev_priv *dpriv = dev->priv;
+
+	unregister_hdlc_device(&dpriv->hdlc);
+	dpriv->usecount--;
+}
+
+static struct pci_device_id dscc4_pci_tbl[] __devinitdata = {
+	{ PCI_VENDOR_ID_SIEMENS, PCI_DEVICE_ID_SIEMENS_DSCC4,
+	        PCI_ANY_ID, PCI_ANY_ID, },
+	{ 0,}
+};
+MODULE_DEVICE_TABLE(pci, dscc4_pci_tbl);
+
+static struct pci_driver dscc4_driver = {
+	name:           "dscc4",
+	id_table:       dscc4_pci_tbl,
+	probe:          dscc4_init_one,
+	remove:         dscc4_remove_one,
+};
+
+static int __init dscc4_init_module(void)
+{
+	return pci_module_init(&dscc4_driver);
+}
+
+static void __exit dscc4_cleanup_module(void)
+{
+	pci_unregister_driver(&dscc4_driver);
+}
+
+module_init(dscc4_init_module);
+module_exit(dscc4_cleanup_module);

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