kernel_optimize_test/drivers/net/ni5010.c

809 lines
23 KiB
C
Raw Normal View History

/* ni5010.c: A network driver for the MiCom-Interlan NI5010 ethercard.
*
* Copyright 1996,1997,2006 Jan-Pascal van Best and Andreas Mohr.
*
* This software may be used and distributed according to the terms
* of the GNU General Public License, incorporated herein by reference.
*
* The authors may be reached as:
* janpascal@vanbest.org andi@lisas.de
*
* Sources:
* Donald Becker's "skeleton.c"
* Crynwr ni5010 packet driver
*
* Changes:
* v0.0: First test version
* v0.1: First working version
* v0.2:
* v0.3->v0.90: Now demand setting io and irq when loading as module
* 970430 v0.91: modified for Linux 2.1.14
* v0.92: Implemented Andreas' (better) NI5010 probe
* 970503 v0.93: Fixed auto-irq failure on warm reboot (JB)
* 970623 v1.00: First kernel version (AM)
* 970814 v1.01: Added detection of onboard receive buffer size (AM)
* 060611 v1.02: slight cleanup: email addresses, driver modernization.
* Bugs:
* - not SMP-safe (no locking of I/O accesses)
* - Note that you have to patch ifconfig for the new /proc/net/dev
* format. It gives incorrect stats otherwise.
*
* To do:
* Fix all bugs :-)
* Move some stuff to chipset_init()
* Handle xmt errors other than collisions
* Complete merge with Andreas' driver
* Implement ring buffers (Is this useful? You can't squeeze
* too many packet in a 2k buffer!)
* Implement DMA (Again, is this useful? Some docs say DMA is
* slower than programmed I/O)
*
* Compile with:
* gcc -O2 -fomit-frame-pointer -m486 -D__KERNEL__ \
* -DMODULE -c ni5010.c
*
* Insert with e.g.:
* insmod ni5010.ko io=0x300 irq=5
*/
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/errno.h>
#include <linux/ioport.h>
#include <linux/slab.h>
#include <linux/interrupt.h>
#include <linux/delay.h>
#include <linux/init.h>
#include <linux/bitops.h>
#include <asm/io.h>
#include <asm/dma.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/skbuff.h>
#include "ni5010.h"
static const char boardname[] = "NI5010";
static char version[] __initdata =
"ni5010.c: v1.02 20060611 Jan-Pascal van Best and Andreas Mohr\n";
/* bufsize_rcv == 0 means autoprobing */
static unsigned int bufsize_rcv;
#define JUMPERED_INTERRUPTS /* IRQ line jumpered on board */
#undef JUMPERED_DMA /* No DMA used */
#undef FULL_IODETECT /* Only detect in portlist */
#ifndef FULL_IODETECT
/* A zero-terminated list of I/O addresses to be probed. */
static unsigned int ports[] __initdata =
{ 0x300, 0x320, 0x340, 0x360, 0x380, 0x3a0, 0 };
#endif
/* Use 0 for production, 1 for verification, >2 for debug */
#ifndef NI5010_DEBUG
#define NI5010_DEBUG 0
#endif
/* Information that needs to be kept for each board. */
struct ni5010_local {
struct net_device_stats stats;
int o_pkt_size;
spinlock_t lock;
};
/* Index to functions, as function prototypes. */
static int ni5010_probe1(struct net_device *dev, int ioaddr);
static int ni5010_open(struct net_device *dev);
static int ni5010_send_packet(struct sk_buff *skb, struct net_device *dev);
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t ni5010_interrupt(int irq, void *dev_id);
static void ni5010_rx(struct net_device *dev);
static void ni5010_timeout(struct net_device *dev);
static int ni5010_close(struct net_device *dev);
static struct net_device_stats *ni5010_get_stats(struct net_device *dev);
static void ni5010_set_multicast_list(struct net_device *dev);
static void reset_receiver(struct net_device *dev);
static int process_xmt_interrupt(struct net_device *dev);
#define tx_done(dev) 1
static void hardware_send_packet(struct net_device *dev, char *buf, int length, int pad);
static void chipset_init(struct net_device *dev, int startp);
static void dump_packet(void *buf, int len);
static void ni5010_show_registers(struct net_device *dev);
static int io;
static int irq;
struct net_device * __init ni5010_probe(int unit)
{
struct net_device *dev = alloc_etherdev(sizeof(struct ni5010_local));
int *port;
int err = 0;
if (!dev)
return ERR_PTR(-ENOMEM);
if (unit >= 0) {
sprintf(dev->name, "eth%d", unit);
netdev_boot_setup_check(dev);
io = dev->base_addr;
irq = dev->irq;
}
PRINTK2((KERN_DEBUG "%s: Entering ni5010_probe\n", dev->name));
SET_MODULE_OWNER(dev);
if (io > 0x1ff) { /* Check a single specified location. */
err = ni5010_probe1(dev, io);
} else if (io != 0) { /* Don't probe at all. */
err = -ENXIO;
} else {
#ifdef FULL_IODETECT
for (io=0x200; io<0x400 && ni5010_probe1(dev, io) ; io+=0x20)
;
if (io == 0x400)
err = -ENODEV;
#else
for (port = ports; *port && ni5010_probe1(dev, *port); port++)
;
if (!*port)
err = -ENODEV;
#endif /* FULL_IODETECT */
}
if (err)
goto out;
err = register_netdev(dev);
if (err)
goto out1;
return dev;
out1:
release_region(dev->base_addr, NI5010_IO_EXTENT);
out:
free_netdev(dev);
return ERR_PTR(err);
}
static inline int rd_port(int ioaddr)
{
inb(IE_RBUF);
return inb(IE_SAPROM);
}
static void __init trigger_irq(int ioaddr)
{
outb(0x00, EDLC_RESET); /* Clear EDLC hold RESET state */
outb(0x00, IE_RESET); /* Board reset */
outb(0x00, EDLC_XMASK); /* Disable all Xmt interrupts */
outb(0x00, EDLC_RMASK); /* Disable all Rcv interrupt */
outb(0xff, EDLC_XCLR); /* Clear all pending Xmt interrupts */
outb(0xff, EDLC_RCLR); /* Clear all pending Rcv interrupts */
/*
* Transmit packet mode: Ignore parity, Power xcvr,
* Enable loopback
*/
outb(XMD_IG_PAR | XMD_T_MODE | XMD_LBC, EDLC_XMODE);
outb(RMD_BROADCAST, EDLC_RMODE); /* Receive normal&broadcast */
outb(XM_ALL, EDLC_XMASK); /* Enable all Xmt interrupts */
udelay(50); /* FIXME: Necessary? */
outb(MM_EN_XMT|MM_MUX, IE_MMODE); /* Start transmission */
}
/*
* This is the real probe routine. Linux has a history of friendly device
* probes on the ISA bus. A good device probes avoids doing writes, and
* verifies that the correct device exists and functions.
*/
static int __init ni5010_probe1(struct net_device *dev, int ioaddr)
{
static unsigned version_printed;
struct ni5010_local *lp;
int i;
unsigned int data = 0;
int boguscount = 40;
int err = -ENODEV;
dev->base_addr = ioaddr;
dev->irq = irq;
if (!request_region(ioaddr, NI5010_IO_EXTENT, boardname))
return -EBUSY;
/*
* This is no "official" probe method, I've rather tested which
* probe works best with my seven NI5010 cards
* (they have very different serial numbers)
* Suggestions or failure reports are very, very welcome !
* But I think it is a relatively good probe method
* since it doesn't use any "outb"
* It should be nearly 100% reliable !
* well-known WARNING: this probe method (like many others)
* will hang the system if a NE2000 card region is probed !
*
* - Andreas
*/
PRINTK2((KERN_DEBUG "%s: entering ni5010_probe1(%#3x)\n",
dev->name, ioaddr));
if (inb(ioaddr+0) == 0xff)
goto out;
while ( (rd_port(ioaddr) & rd_port(ioaddr) & rd_port(ioaddr) &
rd_port(ioaddr) & rd_port(ioaddr) & rd_port(ioaddr)) != 0xff)
{
if (boguscount-- == 0)
goto out;
}
PRINTK2((KERN_DEBUG "%s: I/O #1 passed!\n", dev->name));
for (i=0; i<32; i++)
if ( (data = rd_port(ioaddr)) != 0xff) break;
if (data==0xff)
goto out;
PRINTK2((KERN_DEBUG "%s: I/O #2 passed!\n", dev->name));
if ((data != SA_ADDR0) || (rd_port(ioaddr) != SA_ADDR1) ||
(rd_port(ioaddr) != SA_ADDR2))
goto out;
for (i=0; i<4; i++)
rd_port(ioaddr);
if ( (rd_port(ioaddr) != NI5010_MAGICVAL1) ||
(rd_port(ioaddr) != NI5010_MAGICVAL2) )
goto out;
PRINTK2((KERN_DEBUG "%s: I/O #3 passed!\n", dev->name));
if (NI5010_DEBUG && version_printed++ == 0)
printk(KERN_INFO "%s", version);
printk("NI5010 ethercard probe at 0x%x: ", ioaddr);
dev->base_addr = ioaddr;
for (i=0; i<6; i++) {
outw(i, IE_GP);
printk("%2.2x ", dev->dev_addr[i] = inb(IE_SAPROM));
}
PRINTK2((KERN_DEBUG "%s: I/O #4 passed!\n", dev->name));
#ifdef JUMPERED_INTERRUPTS
if (dev->irq == 0xff)
;
else if (dev->irq < 2) {
unsigned long irq_mask;
PRINTK2((KERN_DEBUG "%s: I/O #5 passed!\n", dev->name));
irq_mask = probe_irq_on();
trigger_irq(ioaddr);
mdelay(20);
dev->irq = probe_irq_off(irq_mask);
PRINTK2((KERN_DEBUG "%s: I/O #6 passed!\n", dev->name));
if (dev->irq == 0) {
err = -EAGAIN;
printk(KERN_WARNING "%s: no IRQ found!\n", dev->name);
goto out;
}
PRINTK2((KERN_DEBUG "%s: I/O #7 passed!\n", dev->name));
} else if (dev->irq == 2) {
dev->irq = 9;
}
#endif /* JUMPERED_INTERRUPTS */
PRINTK2((KERN_DEBUG "%s: I/O #9 passed!\n", dev->name));
/* DMA is not supported (yet?), so no use detecting it */
lp = netdev_priv(dev);
spin_lock_init(&lp->lock);
PRINTK2((KERN_DEBUG "%s: I/O #10 passed!\n", dev->name));
/* get the size of the onboard receive buffer
* higher addresses than bufsize are wrapped into real buffer
* i.e. data for offs. 0x801 is written to 0x1 with a 2K onboard buffer
*/
if (!bufsize_rcv) {
outb(1, IE_MMODE); /* Put Rcv buffer on system bus */
outw(0, IE_GP); /* Point GP at start of packet */
outb(0, IE_RBUF); /* set buffer byte 0 to 0 */
for (i = 1; i < 0xff; i++) {
outw(i << 8, IE_GP); /* Point GP at packet size to be tested */
outb(i, IE_RBUF);
outw(0x0, IE_GP); /* Point GP at start of packet */
data = inb(IE_RBUF);
if (data == i) break;
}
bufsize_rcv = i << 8;
outw(0, IE_GP); /* Point GP at start of packet */
outb(0, IE_RBUF); /* set buffer byte 0 to 0 again */
}
printk("-> bufsize rcv/xmt=%d/%d\n", bufsize_rcv, NI5010_BUFSIZE);
memset(dev->priv, 0, sizeof(struct ni5010_local));
dev->open = ni5010_open;
dev->stop = ni5010_close;
dev->hard_start_xmit = ni5010_send_packet;
dev->get_stats = ni5010_get_stats;
dev->set_multicast_list = ni5010_set_multicast_list;
dev->tx_timeout = ni5010_timeout;
dev->watchdog_timeo = HZ/20;
dev->flags &= ~IFF_MULTICAST; /* Multicast doesn't work */
/* Shut up the ni5010 */
outb(0, EDLC_RMASK); /* Mask all receive interrupts */
outb(0, EDLC_XMASK); /* Mask all xmit interrupts */
outb(0xff, EDLC_RCLR); /* Kill all pending rcv interrupts */
outb(0xff, EDLC_XCLR); /* Kill all pending xmt interrupts */
printk(KERN_INFO "%s: NI5010 found at 0x%x, using IRQ %d", dev->name, ioaddr, dev->irq);
if (dev->dma)
printk(" & DMA %d", dev->dma);
printk(".\n");
return 0;
out:
release_region(dev->base_addr, NI5010_IO_EXTENT);
return err;
}
/*
* 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 a non-reboot way to recover if something goes wrong.
*/
static int ni5010_open(struct net_device *dev)
{
int ioaddr = dev->base_addr;
int i;
PRINTK2((KERN_DEBUG "%s: entering ni5010_open()\n", dev->name));
if (request_irq(dev->irq, &ni5010_interrupt, 0, boardname, dev)) {
printk(KERN_WARNING "%s: Cannot get irq %#2x\n", dev->name, dev->irq);
return -EAGAIN;
}
PRINTK3((KERN_DEBUG "%s: passed open() #1\n", dev->name));
/*
* Always allocate the DMA channel after the IRQ,
* and clean up on failure.
*/
#ifdef JUMPERED_DMA
if (request_dma(dev->dma, cardname)) {
printk(KERN_WARNING "%s: Cannot get dma %#2x\n", dev->name, dev->dma);
free_irq(dev->irq, NULL);
return -EAGAIN;
}
#endif /* JUMPERED_DMA */
PRINTK3((KERN_DEBUG "%s: passed open() #2\n", dev->name));
/* Reset the hardware here. Don't forget to set the station address. */
outb(RS_RESET, EDLC_RESET); /* Hold up EDLC_RESET while configing board */
outb(0, IE_RESET); /* Hardware reset of ni5010 board */
outb(XMD_LBC, EDLC_XMODE); /* Only loopback xmits */
PRINTK3((KERN_DEBUG "%s: passed open() #3\n", dev->name));
/* Set the station address */
for(i = 0;i < 6; i++) {
outb(dev->dev_addr[i], EDLC_ADDR + i);
}
PRINTK3((KERN_DEBUG "%s: Initialising ni5010\n", dev->name));
outb(0, EDLC_XMASK); /* No xmit interrupts for now */
outb(XMD_IG_PAR | XMD_T_MODE | XMD_LBC, EDLC_XMODE);
/* Normal packet xmit mode */
outb(0xff, EDLC_XCLR); /* Clear all pending xmit interrupts */
outb(RMD_BROADCAST, EDLC_RMODE);
/* Receive broadcast and normal packets */
reset_receiver(dev); /* Ready ni5010 for receiving packets */
outb(0, EDLC_RESET); /* Un-reset the ni5010 */
netif_start_queue(dev);
if (NI5010_DEBUG) ni5010_show_registers(dev);
PRINTK((KERN_DEBUG "%s: open successful\n", dev->name));
return 0;
}
static void reset_receiver(struct net_device *dev)
{
int ioaddr = dev->base_addr;
PRINTK3((KERN_DEBUG "%s: resetting receiver\n", dev->name));
outw(0, IE_GP); /* Receive packet at start of buffer */
outb(0xff, EDLC_RCLR); /* Clear all pending rcv interrupts */
outb(0, IE_MMODE); /* Put EDLC to rcv buffer */
outb(MM_EN_RCV, IE_MMODE); /* Enable rcv */
outb(0xff, EDLC_RMASK); /* Enable all rcv interrupts */
}
static void ni5010_timeout(struct net_device *dev)
{
printk(KERN_WARNING "%s: transmit timed out, %s?\n", dev->name,
tx_done(dev) ? "IRQ conflict" : "network cable problem");
/* Try to restart the adaptor. */
/* FIXME: Give it a real kick here */
chipset_init(dev, 1);
dev->trans_start = jiffies;
netif_wake_queue(dev);
}
static int ni5010_send_packet(struct sk_buff *skb, struct net_device *dev)
{
int length = ETH_ZLEN < skb->len ? skb->len : ETH_ZLEN;
PRINTK2((KERN_DEBUG "%s: entering ni5010_send_packet\n", dev->name));
/*
* Block sending
*/
netif_stop_queue(dev);
hardware_send_packet(dev, (unsigned char *)skb->data, skb->len, length-skb->len);
dev->trans_start = jiffies;
dev_kfree_skb (skb);
return 0;
}
/*
* The typical workload of the driver:
* Handle the network interface interrupts.
*/
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t ni5010_interrupt(int irq, void *dev_id)
{
struct net_device *dev = dev_id;
struct ni5010_local *lp;
int ioaddr, status;
int xmit_was_error = 0;
PRINTK2((KERN_DEBUG "%s: entering ni5010_interrupt\n", dev->name));
ioaddr = dev->base_addr;
lp = netdev_priv(dev);
spin_lock(&lp->lock);
status = inb(IE_ISTAT);
PRINTK3((KERN_DEBUG "%s: IE_ISTAT = %#02x\n", dev->name, status));
if ((status & IS_R_INT) == 0) ni5010_rx(dev);
if ((status & IS_X_INT) == 0) {
xmit_was_error = process_xmt_interrupt(dev);
}
if ((status & IS_DMA_INT) == 0) {
PRINTK((KERN_DEBUG "%s: DMA complete (?)\n", dev->name));
outb(0, IE_DMA_RST); /* Reset DMA int */
}
if (!xmit_was_error)
reset_receiver(dev);
spin_unlock(&lp->lock);
return IRQ_HANDLED;
}
static void dump_packet(void *buf, int len)
{
int i;
printk(KERN_DEBUG "Packet length = %#4x\n", len);
for (i = 0; i < len; i++){
if (i % 16 == 0) printk(KERN_DEBUG "%#4.4x", i);
if (i % 2 == 0) printk(" ");
printk("%2.2x", ((unsigned char *)buf)[i]);
if (i % 16 == 15) printk("\n");
}
printk("\n");
return;
}
/* We have a good packet, get it out of the buffer. */
static void ni5010_rx(struct net_device *dev)
{
struct ni5010_local *lp = netdev_priv(dev);
int ioaddr = dev->base_addr;
unsigned char rcv_stat;
struct sk_buff *skb;
int i_pkt_size;
PRINTK2((KERN_DEBUG "%s: entering ni5010_rx()\n", dev->name));
rcv_stat = inb(EDLC_RSTAT);
PRINTK3((KERN_DEBUG "%s: EDLC_RSTAT = %#2x\n", dev->name, rcv_stat));
if ( (rcv_stat & RS_VALID_BITS) != RS_PKT_OK) {
PRINTK((KERN_INFO "%s: receive error.\n", dev->name));
lp->stats.rx_errors++;
if (rcv_stat & RS_RUNT) lp->stats.rx_length_errors++;
if (rcv_stat & RS_ALIGN) lp->stats.rx_frame_errors++;
if (rcv_stat & RS_CRC_ERR) lp->stats.rx_crc_errors++;
if (rcv_stat & RS_OFLW) lp->stats.rx_fifo_errors++;
outb(0xff, EDLC_RCLR); /* Clear the interrupt */
return;
}
outb(0xff, EDLC_RCLR); /* Clear the interrupt */
i_pkt_size = inw(IE_RCNT);
if (i_pkt_size > ETH_FRAME_LEN || i_pkt_size < 10 ) {
PRINTK((KERN_DEBUG "%s: Packet size error, packet size = %#4.4x\n",
dev->name, i_pkt_size));
lp->stats.rx_errors++;
lp->stats.rx_length_errors++;
return;
}
/* Malloc up new buffer. */
skb = dev_alloc_skb(i_pkt_size + 3);
if (skb == NULL) {
printk(KERN_WARNING "%s: Memory squeeze, dropping packet.\n", dev->name);
lp->stats.rx_dropped++;
return;
}
skb->dev = dev;
skb_reserve(skb, 2);
/* Read packet into buffer */
outb(MM_MUX, IE_MMODE); /* Rcv buffer to system bus */
outw(0, IE_GP); /* Seek to beginning of packet */
insb(IE_RBUF, skb_put(skb, i_pkt_size), i_pkt_size);
if (NI5010_DEBUG >= 4)
dump_packet(skb->data, skb->len);
skb->protocol = eth_type_trans(skb,dev);
netif_rx(skb);
dev->last_rx = jiffies;
lp->stats.rx_packets++;
lp->stats.rx_bytes += i_pkt_size;
PRINTK2((KERN_DEBUG "%s: Received packet, size=%#4.4x\n",
dev->name, i_pkt_size));
}
static int process_xmt_interrupt(struct net_device *dev)
{
struct ni5010_local *lp = netdev_priv(dev);
int ioaddr = dev->base_addr;
int xmit_stat;
PRINTK2((KERN_DEBUG "%s: entering process_xmt_interrupt\n", dev->name));
xmit_stat = inb(EDLC_XSTAT);
PRINTK3((KERN_DEBUG "%s: EDLC_XSTAT = %2.2x\n", dev->name, xmit_stat));
outb(0, EDLC_XMASK); /* Disable xmit IRQ's */
outb(0xff, EDLC_XCLR); /* Clear all pending xmit IRQ's */
if (xmit_stat & XS_COLL){
PRINTK((KERN_DEBUG "%s: collision detected, retransmitting\n",
dev->name));
outw(NI5010_BUFSIZE - lp->o_pkt_size, IE_GP);
/* outb(0, IE_MMODE); */ /* xmt buf on sysbus FIXME: needed ? */
outb(MM_EN_XMT | MM_MUX, IE_MMODE);
outb(XM_ALL, EDLC_XMASK); /* Enable xmt IRQ's */
lp->stats.collisions++;
return 1;
}
/* FIXME: handle other xmt error conditions */
lp->stats.tx_packets++;
lp->stats.tx_bytes += lp->o_pkt_size;
netif_wake_queue(dev);
PRINTK2((KERN_DEBUG "%s: sent packet, size=%#4.4x\n",
dev->name, lp->o_pkt_size));
return 0;
}
/* The inverse routine to ni5010_open(). */
static int ni5010_close(struct net_device *dev)
{
int ioaddr = dev->base_addr;
PRINTK2((KERN_DEBUG "%s: entering ni5010_close\n", dev->name));
#ifdef JUMPERED_INTERRUPTS
free_irq(dev->irq, NULL);
#endif
/* Put card in held-RESET state */
outb(0, IE_MMODE);
outb(RS_RESET, EDLC_RESET);
netif_stop_queue(dev);
PRINTK((KERN_DEBUG "%s: %s closed down\n", dev->name, boardname));
return 0;
}
/* Get the current statistics. This may be called with the card open or
closed. */
static struct net_device_stats *ni5010_get_stats(struct net_device *dev)
{
struct ni5010_local *lp = netdev_priv(dev);
PRINTK2((KERN_DEBUG "%s: entering ni5010_get_stats\n", dev->name));
if (NI5010_DEBUG) ni5010_show_registers(dev);
/* cli(); */
/* Update the statistics from the device registers. */
/* We do this in the interrupt handler */
/* sti(); */
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 ni5010_set_multicast_list(struct net_device *dev)
{
short ioaddr = dev->base_addr;
PRINTK2((KERN_DEBUG "%s: entering set_multicast_list\n", dev->name));
if (dev->flags&IFF_PROMISC || dev->flags&IFF_ALLMULTI) {
dev->flags |= IFF_PROMISC;
outb(RMD_PROMISC, EDLC_RMODE); /* Enable promiscuous mode */
PRINTK((KERN_DEBUG "%s: Entering promiscuous mode\n", dev->name));
} else if (dev->mc_list) {
/* Sorry, multicast not supported */
PRINTK((KERN_DEBUG "%s: No multicast, entering broadcast mode\n", dev->name));
outb(RMD_BROADCAST, EDLC_RMODE);
} else {
PRINTK((KERN_DEBUG "%s: Entering broadcast mode\n", dev->name));
outb(RMD_BROADCAST, EDLC_RMODE); /* Disable promiscuous mode, use normal mode */
}
}
static void hardware_send_packet(struct net_device *dev, char *buf, int length, int pad)
{
struct ni5010_local *lp = netdev_priv(dev);
int ioaddr = dev->base_addr;
unsigned long flags;
unsigned int buf_offs;
PRINTK2((KERN_DEBUG "%s: entering hardware_send_packet\n", dev->name));
if (length > ETH_FRAME_LEN) {
PRINTK((KERN_WARNING "%s: packet too large, not possible\n",
dev->name));
return;
}
if (NI5010_DEBUG) ni5010_show_registers(dev);
if (inb(IE_ISTAT) & IS_EN_XMT) {
PRINTK((KERN_WARNING "%s: sending packet while already transmitting, not possible\n",
dev->name));
return;
}
if (NI5010_DEBUG > 3) dump_packet(buf, length);
buf_offs = NI5010_BUFSIZE - length - pad;
spin_lock_irqsave(&lp->lock, flags);
lp->o_pkt_size = length + pad;
outb(0, EDLC_RMASK); /* Mask all receive interrupts */
outb(0, IE_MMODE); /* Put Xmit buffer on system bus */
outb(0xff, EDLC_RCLR); /* Clear out pending rcv interrupts */
outw(buf_offs, IE_GP); /* Point GP at start of packet */
outsb(IE_XBUF, buf, length); /* Put data in buffer */
while(pad--)
outb(0, IE_XBUF);
outw(buf_offs, IE_GP); /* Rewrite where packet starts */
/* should work without that outb() (Crynwr used it) */
/*outb(MM_MUX, IE_MMODE);*/ /* Xmt buffer to EDLC bus */
outb(MM_EN_XMT | MM_MUX, IE_MMODE); /* Begin transmission */
outb(XM_ALL, EDLC_XMASK); /* Cause interrupt after completion or fail */
spin_unlock_irqrestore(&lp->lock, flags);
netif_wake_queue(dev);
if (NI5010_DEBUG) ni5010_show_registers(dev);
}
static void chipset_init(struct net_device *dev, int startp)
{
/* FIXME: Move some stuff here */
PRINTK3((KERN_DEBUG "%s: doing NOTHING in chipset_init\n", dev->name));
}
static void ni5010_show_registers(struct net_device *dev)
{
int ioaddr = dev->base_addr;
PRINTK3((KERN_DEBUG "%s: XSTAT %#2.2x\n", dev->name, inb(EDLC_XSTAT)));
PRINTK3((KERN_DEBUG "%s: XMASK %#2.2x\n", dev->name, inb(EDLC_XMASK)));
PRINTK3((KERN_DEBUG "%s: RSTAT %#2.2x\n", dev->name, inb(EDLC_RSTAT)));
PRINTK3((KERN_DEBUG "%s: RMASK %#2.2x\n", dev->name, inb(EDLC_RMASK)));
PRINTK3((KERN_DEBUG "%s: RMODE %#2.2x\n", dev->name, inb(EDLC_RMODE)));
PRINTK3((KERN_DEBUG "%s: XMODE %#2.2x\n", dev->name, inb(EDLC_XMODE)));
PRINTK3((KERN_DEBUG "%s: ISTAT %#2.2x\n", dev->name, inb(IE_ISTAT)));
}
#ifdef MODULE
static struct net_device *dev_ni5010;
module_param(io, int, 0);
module_param(irq, int, 0);
MODULE_PARM_DESC(io, "ni5010 I/O base address");
MODULE_PARM_DESC(irq, "ni5010 IRQ number");
static int __init ni5010_init_module(void)
{
PRINTK2((KERN_DEBUG "%s: entering init_module\n", boardname));
/*
if(io <= 0 || irq == 0){
printk(KERN_WARNING "%s: Autoprobing not allowed for modules.\n", boardname);
printk(KERN_WARNING "%s: Set symbols 'io' and 'irq'\n", boardname);
return -EINVAL;
}
*/
if (io <= 0){
printk(KERN_WARNING "%s: Autoprobing for modules is hazardous, trying anyway..\n", boardname);
}
PRINTK2((KERN_DEBUG "%s: init_module irq=%#2x, io=%#3x\n", boardname, irq, io));
dev_ni5010 = ni5010_probe(-1);
if (IS_ERR(dev_ni5010))
return PTR_ERR(dev_ni5010);
return 0;
}
static void __exit ni5010_cleanup_module(void)
{
PRINTK2((KERN_DEBUG "%s: entering cleanup_module\n", boardname));
unregister_netdev(dev_ni5010);
release_region(dev_ni5010->base_addr, NI5010_IO_EXTENT);
free_netdev(dev_ni5010);
}
module_init(ni5010_init_module);
module_exit(ni5010_cleanup_module);
#endif /* MODULE */
MODULE_LICENSE("GPL");
/*
* Local variables:
* compile-command: "gcc -D__KERNEL__ -I/usr/src/linux/net/inet -Wall -Wstrict-prototypes -O6 -m486 -c ni5010.c"
* version-control: t
* kept-new-versions: 5
* tab-width: 4
* End:
*/