UbiquitousOS
/
OpenCloudOS-Kernel
mirror of https://gitee.com/OpenCloudOS/OpenCloudOS-Kernel.git
11
0
Fork 0
Code Issues Projects Releases Wiki Activity
OpenCloudOS-Kernel/drivers/net/ethernet/seeq/ether3.c

900 lines
23 KiB
C
Raw Normal View History

Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/*
* linux/drivers/acorn/net/ether3.c
*
* Copyright (C) 1995-2000 Russell King
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*
* SEEQ nq8005 ethernet driver for Acorn/ANT Ether3 card
* for Acorn machines
*
* By Russell King, with some suggestions from borris@ant.co.uk
*
* Changelog:
* 1.04 RMK 29/02/1996 Won't pass packets that are from our ethernet
* address up to the higher levels - they're
* silently ignored. I/F can now be put into
* multicast mode. Receiver routine optimised.
* 1.05 RMK 30/02/1996 Now claims interrupt at open when part of
* the kernel rather than when a module.
* 1.06 RMK 02/03/1996 Various code cleanups
* 1.07 RMK 13/10/1996 Optimised interrupt routine and transmit
* routines.
* 1.08 RMK 14/10/1996 Fixed problem with too many packets,
* prevented the kernel message about dropped
* packets appearing too many times a second.
* Now does not disable all IRQs, only the IRQ
* used by this card.
* 1.09 RMK 10/11/1996 Only enables TX irq when buffer space is low,
* but we still service the TX queue if we get a
* RX interrupt.
* 1.10 RMK 15/07/1997 Fixed autoprobing of NQ8004.
* 1.11 RMK 16/11/1997 Fixed autoprobing of NQ8005A.
* 1.12 RMK 31/12/1997 Removed reference to dev_tint for Linux 2.1.
* RMK 27/06/1998 Changed asm/delay.h to linux/delay.h.
* 1.13 RMK 29/06/1998 Fixed problem with transmission of packets.
* Chip seems to have a bug in, whereby if the
* packet starts two bytes from the end of the
* buffer, it corrupts the receiver chain, and
* never updates the transmit status correctly.
* 1.14 RMK 07/01/1998 Added initial code for ETHERB addressing.
* 1.15 RMK 30/04/1999 More fixes to the transmit routine for buggy
* hardware.
* 1.16 RMK 10/02/2000 Updated for 2.3.43
* 1.17 RMK 13/05/2000 Updated for 2.3.99-pre8
*/
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/types.h>
#include <linux/fcntl.h>
#include <linux/interrupt.h>
#include <linux/ioport.h>
#include <linux/in.h>
#include <linux/slab.h>
#include <linux/string.h>
#include <linux/errno.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/skbuff.h>
#include <linux/device.h>
#include <linux/init.h>
#include <linux/delay.h>
#include <linux/bitops.h>
#include <asm/ecard.h>
#include <asm/io.h>
seeq: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Russell King <linux@arm.linux.org.uk> Cc: linux-arm-kernel@lists.infradead.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 22:23:31 +08:00
static char version[] = "ether3 ethernet driver (c) 1995-2000 R.M.King v1.17\n";
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#include "ether3.h"
static unsigned int net_debug = NET_DEBUG;
static void ether3_setmulticastlist(struct net_device *dev);
static int ether3_rx(struct net_device *dev, unsigned int maxcnt);
static void ether3_tx(struct net_device *dev);
static int ether3_open (struct net_device *dev);
static int ether3_sendpacket (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 ether3_interrupt (int irq, void *dev_id);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
static int ether3_close (struct net_device *dev);
static void ether3_setmulticastlist (struct net_device *dev);
static void ether3_timeout(struct net_device *dev);
#define BUS_16 2
#define BUS_8 1
#define BUS_UNKNOWN 0
/* --------------------------------------------------------------------------- */
typedef enum {
buffer_write,
buffer_read
} buffer_rw_t;
/*
* ether3 read/write. Slow things down a bit...
* The SEEQ8005 doesn't like us writing to its registers
* too quickly.
*/
net: seeq: use __iomem pointers for MMIO ARM is moving to stricter checks on readl/write functions, so we need to use the correct types everywhere. Cc: netdev@vger.kernel.org Cc: "David S. Miller" <davem@davemloft.net> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2012-09-15 04:34:35 +08:00
static inline void ether3_outb(int v, void __iomem *r)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
writeb(v, r);
udelay(1);
}
net: seeq: use __iomem pointers for MMIO ARM is moving to stricter checks on readl/write functions, so we need to use the correct types everywhere. Cc: netdev@vger.kernel.org Cc: "David S. Miller" <davem@davemloft.net> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2012-09-15 04:34:35 +08:00
static inline void ether3_outw(int v, void __iomem *r)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
writew(v, r);
udelay(1);
}
#define ether3_inb(r) ({ unsigned int __v = readb((r)); udelay(1); __v; })
#define ether3_inw(r) ({ unsigned int __v = readw((r)); udelay(1); __v; })
static int
ether3_setbuffer(struct net_device *dev, buffer_rw_t read, int start)
{
int timeout = 1000;
ether3_outw(priv(dev)->regs.config1 | CFG1_LOCBUFMEM, REG_CONFIG1);
ether3_outw(priv(dev)->regs.command | CMD_FIFOWRITE, REG_COMMAND);
while ((ether3_inw(REG_STATUS) & STAT_FIFOEMPTY) == 0) {
if (!timeout--) {
printk("%s: setbuffer broken\n", dev->name);
priv(dev)->broken = 1;
return 1;
}
udelay(1);
}
if (read == buffer_read) {
ether3_outw(start, REG_DMAADDR);
ether3_outw(priv(dev)->regs.command | CMD_FIFOREAD, REG_COMMAND);
} else {
ether3_outw(priv(dev)->regs.command | CMD_FIFOWRITE, REG_COMMAND);
ether3_outw(start, REG_DMAADDR);
}
return 0;
}
/*
* write data to the buffer memory
*/
#define ether3_writebuffer(dev,data,length) \
writesw(REG_BUFWIN, (data), (length) >> 1)
#define ether3_writeword(dev,data) \
writew((data), REG_BUFWIN)
#define ether3_writelong(dev,data) { \
void __iomem *reg_bufwin = REG_BUFWIN; \
writew((data), reg_bufwin); \
writew((data) >> 16, reg_bufwin); \
}
/*
* read data from the buffer memory
*/
#define ether3_readbuffer(dev,data,length) \
readsw(REG_BUFWIN, (data), (length) >> 1)
#define ether3_readword(dev) \
readw(REG_BUFWIN)
#define ether3_readlong(dev) \
readw(REG_BUFWIN) | (readw(REG_BUFWIN) << 16)
/*
* Switch LED off...
*/
static void ether3_ledoff(unsigned long data)
{
struct net_device *dev = (struct net_device *)data;
ether3_outw(priv(dev)->regs.config2 |= CFG2_CTRLO, REG_CONFIG2);
}
/*
* switch LED on...
*/
static inline void ether3_ledon(struct net_device *dev)
{
del_timer(&priv(dev)->timer);
priv(dev)->timer.expires = jiffies + HZ / 50; /* leave on for 1/50th second */
priv(dev)->timer.data = (unsigned long)dev;
priv(dev)->timer.function = ether3_ledoff;
add_timer(&priv(dev)->timer);
if (priv(dev)->regs.config2 & CFG2_CTRLO)
ether3_outw(priv(dev)->regs.config2 &= ~CFG2_CTRLO, REG_CONFIG2);
}
/*
* Read the ethernet address string from the on board rom.
* This is an ascii string!!!
*/
seeq: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Russell King <linux@arm.linux.org.uk> Cc: linux-arm-kernel@lists.infradead.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 22:23:31 +08:00
static int
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ether3_addr(char *addr, struct expansion_card *ec)
{
struct in_chunk_dir cd;
char *s;
if (ecard_readchunk(&cd, ec, 0xf5, 0) && (s = strchr(cd.d.string, '('))) {
int i;
for (i = 0; i<6; i++) {
addr[i] = simple_strtoul(s + 1, &s, 0x10);
if (*s != (i==5?')' : ':' ))
break;
}
if (i == 6)
return 0;
}
/* I wonder if we should even let the user continue in this case
* - no, it would be better to disable the device
*/
printk(KERN_ERR "ether3: Couldn't read a valid MAC address from card.\n");
return -ENODEV;
}
/* --------------------------------------------------------------------------- */
seeq: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Russell King <linux@arm.linux.org.uk> Cc: linux-arm-kernel@lists.infradead.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 22:23:31 +08:00
static int
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ether3_ramtest(struct net_device *dev, unsigned char byte)
{
unsigned char *buffer = kmalloc(RX_END, GFP_KERNEL);
int i,ret = 0;
int max_errors = 4;
int bad = -1;
if (!buffer)
return 1;
memset(buffer, byte, RX_END);
ether3_setbuffer(dev, buffer_write, 0);
ether3_writebuffer(dev, buffer, TX_END);
ether3_setbuffer(dev, buffer_write, RX_START);
ether3_writebuffer(dev, buffer + RX_START, RX_LEN);
memset(buffer, byte ^ 0xff, RX_END);
ether3_setbuffer(dev, buffer_read, 0);
ether3_readbuffer(dev, buffer, TX_END);
ether3_setbuffer(dev, buffer_read, RX_START);
ether3_readbuffer(dev, buffer + RX_START, RX_LEN);
for (i = 0; i < RX_END; i++) {
if (buffer[i] != byte) {
if (max_errors > 0 && bad != buffer[i]) {
printk("%s: RAM failed with (%02X instead of %02X) at 0x%04X",
dev->name, buffer[i], byte, i);
ret = 2;
max_errors--;
bad = i;
}
} else {
if (bad != -1) {
if (bad != i - 1)
printk(" - 0x%04X\n", i - 1);
printk("\n");
bad = -1;
}
}
}
if (bad != -1)
printk(" - 0xffff\n");
kfree(buffer);
return ret;
}
/* ------------------------------------------------------------------------------- */
seeq: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Russell King <linux@arm.linux.org.uk> Cc: linux-arm-kernel@lists.infradead.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 22:23:31 +08:00
static int ether3_init_2(struct net_device *dev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
int i;
priv(dev)->regs.config1 = CFG1_RECVCOMPSTAT0|CFG1_DMABURST8;
priv(dev)->regs.config2 = CFG2_CTRLO|CFG2_RECVCRC|CFG2_ERRENCRC;
priv(dev)->regs.command = 0;
/*
* Set up our hardware address
*/
ether3_outw(priv(dev)->regs.config1 | CFG1_BUFSELSTAT0, REG_CONFIG1);
for (i = 0; i < 6; i++)
ether3_outb(dev->dev_addr[i], REG_BUFWIN);
if (dev->flags & IFF_PROMISC)
priv(dev)->regs.config1 |= CFG1_RECVPROMISC;
else if (dev->flags & IFF_MULTICAST)
priv(dev)->regs.config1 |= CFG1_RECVSPECBRMULTI;
else
priv(dev)->regs.config1 |= CFG1_RECVSPECBROAD;
/*
* There is a problem with the NQ8005 in that it occasionally loses the
* last two bytes. To get round this problem, we receive the CRC as
* well. That way, if we do lose the last two, then it doesn't matter.
*/
ether3_outw(priv(dev)->regs.config1 | CFG1_TRANSEND, REG_CONFIG1);
ether3_outw((TX_END>>8) - 1, REG_BUFWIN);
ether3_outw(priv(dev)->rx_head, REG_RECVPTR);
ether3_outw(0, REG_TRANSMITPTR);
ether3_outw(priv(dev)->rx_head >> 8, REG_RECVEND);
ether3_outw(priv(dev)->regs.config2, REG_CONFIG2);
ether3_outw(priv(dev)->regs.config1 | CFG1_LOCBUFMEM, REG_CONFIG1);
ether3_outw(priv(dev)->regs.command, REG_COMMAND);
i = ether3_ramtest(dev, 0x5A);
if(i)
return i;
i = ether3_ramtest(dev, 0x1E);
if(i)
return i;
ether3_setbuffer(dev, buffer_write, 0);
ether3_writelong(dev, 0);
return 0;
}
static void
ether3_init_for_open(struct net_device *dev)
{
int i;
/* Reset the chip */
ether3_outw(CFG2_RESET, REG_CONFIG2);
udelay(4);
priv(dev)->regs.command = 0;
ether3_outw(CMD_RXOFF|CMD_TXOFF, REG_COMMAND);
while (ether3_inw(REG_STATUS) & (STAT_RXON|STAT_TXON))
barrier();
ether3_outw(priv(dev)->regs.config1 | CFG1_BUFSELSTAT0, REG_CONFIG1);
for (i = 0; i < 6; i++)
ether3_outb(dev->dev_addr[i], REG_BUFWIN);
priv(dev)->tx_head = 0;
priv(dev)->tx_tail = 0;
priv(dev)->regs.config2 |= CFG2_CTRLO;
priv(dev)->rx_head = RX_START;
ether3_outw(priv(dev)->regs.config1 | CFG1_TRANSEND, REG_CONFIG1);
ether3_outw((TX_END>>8) - 1, REG_BUFWIN);
ether3_outw(priv(dev)->rx_head, REG_RECVPTR);
ether3_outw(priv(dev)->rx_head >> 8, REG_RECVEND);
ether3_outw(0, REG_TRANSMITPTR);
ether3_outw(priv(dev)->regs.config2, REG_CONFIG2);
ether3_outw(priv(dev)->regs.config1 | CFG1_LOCBUFMEM, REG_CONFIG1);
ether3_setbuffer(dev, buffer_write, 0);
ether3_writelong(dev, 0);
priv(dev)->regs.command = CMD_ENINTRX | CMD_ENINTTX;
ether3_outw(priv(dev)->regs.command | CMD_RXON, REG_COMMAND);
}
static inline int
ether3_probe_bus_8(struct net_device *dev, int val)
{
int write_low, write_high, read_low, read_high;
write_low = val & 255;
write_high = val >> 8;
printk(KERN_DEBUG "ether3_probe: write8 [%02X:%02X]", write_high, write_low);
ether3_outb(write_low, REG_RECVPTR);
ether3_outb(write_high, REG_RECVPTR + 4);
read_low = ether3_inb(REG_RECVPTR);
read_high = ether3_inb(REG_RECVPTR + 4);
printk(", read8 [%02X:%02X]\n", read_high, read_low);
return read_low == write_low && read_high == write_high;
}
static inline int
ether3_probe_bus_16(struct net_device *dev, int val)
{
int read_val;
ether3_outw(val, REG_RECVPTR);
read_val = ether3_inw(REG_RECVPTR);
printk(KERN_DEBUG "ether3_probe: write16 [%04X], read16 [%04X]\n", val, read_val);
return read_val == val;
}
/*
* 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
ether3_open(struct net_device *dev)
{
if (request_irq(dev->irq, ether3_interrupt, 0, "ether3", dev))
return -EAGAIN;
ether3_init_for_open(dev);
netif_start_queue(dev);
return 0;
}
/*
* The inverse routine to ether3_open().
*/
static int
ether3_close(struct net_device *dev)
{
netif_stop_queue(dev);
disable_irq(dev->irq);
ether3_outw(CMD_RXOFF|CMD_TXOFF, REG_COMMAND);
priv(dev)->regs.command = 0;
while (ether3_inw(REG_STATUS) & (STAT_RXON|STAT_TXON))
barrier();
ether3_outb(0x80, REG_CONFIG2 + 4);
ether3_outw(0, REG_COMMAND);
free_irq(dev->irq, dev);
return 0;
}
/*
* Set or clear promiscuous/multicast mode filter for this adaptor.
*
* We don't attempt any packet filtering. The card may have a SEEQ 8004
* in which does not have the other ethernet address registers present...
*/
static void ether3_setmulticastlist(struct net_device *dev)
{
priv(dev)->regs.config1 &= ~CFG1_RECVPROMISC;
if (dev->flags & IFF_PROMISC) {
/* promiscuous mode */
priv(dev)->regs.config1 |= CFG1_RECVPROMISC;
net: use netdev_mc_count and netdev_mc_empty when appropriate This patch replaces dev->mc_count in all drivers (hopefully I didn't miss anything). Used spatch and did small tweaks and conding style changes when it was suitable. Jirka Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-02-08 12:30:35 +08:00
} else if (dev->flags & IFF_ALLMULTI || !netdev_mc_empty(dev)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
priv(dev)->regs.config1 |= CFG1_RECVSPECBRMULTI;
} else
priv(dev)->regs.config1 |= CFG1_RECVSPECBROAD;
ether3_outw(priv(dev)->regs.config1 | CFG1_LOCBUFMEM, REG_CONFIG1);
}
static void ether3_timeout(struct net_device *dev)
{
unsigned long flags;
del_timer(&priv(dev)->timer);
local_irq_save(flags);
printk(KERN_ERR "%s: transmit timed out, network cable problem?\n", dev->name);
printk(KERN_ERR "%s: state: { status=%04X cfg1=%04X cfg2=%04X }\n", dev->name,
ether3_inw(REG_STATUS), ether3_inw(REG_CONFIG1), ether3_inw(REG_CONFIG2));
printk(KERN_ERR "%s: { rpr=%04X rea=%04X tpr=%04X }\n", dev->name,
ether3_inw(REG_RECVPTR), ether3_inw(REG_RECVEND), ether3_inw(REG_TRANSMITPTR));
printk(KERN_ERR "%s: tx head=%X tx tail=%X\n", dev->name,
priv(dev)->tx_head, priv(dev)->tx_tail);
ether3_setbuffer(dev, buffer_read, priv(dev)->tx_tail);
printk(KERN_ERR "%s: packet status = %08X\n", dev->name, ether3_readlong(dev));
local_irq_restore(flags);
priv(dev)->regs.config2 |= CFG2_CTRLO;
ether3: Use net_device_stats from struct net_device struct net_device has its own struct net_device_stats member, so use this one instead of a private copy in the ether1_priv struct. As the new ndo_get_stats function would just return dev->stats we can omit it. This patch also removes an incorrect memset of the stats on open. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 04:15:31 +08:00
dev->stats.tx_errors += 1;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ether3_outw(priv(dev)->regs.config2, REG_CONFIG2);
priv(dev)->tx_head = priv(dev)->tx_tail = 0;
netif_wake_queue(dev);
}
/*
* Transmit a packet
*/
static int
ether3_sendpacket(struct sk_buff *skb, struct net_device *dev)
{
unsigned long flags;
unsigned int length = ETH_ZLEN < skb->len ? skb->len : ETH_ZLEN;
unsigned int ptr, next_ptr;
if (priv(dev)->broken) {
dev_kfree_skb(skb);
ether3: Use net_device_stats from struct net_device struct net_device has its own struct net_device_stats member, so use this one instead of a private copy in the ether1_priv struct. As the new ndo_get_stats function would just return dev->stats we can omit it. This patch also removes an incorrect memset of the stats on open. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 04:15:31 +08:00
dev->stats.tx_dropped++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
netif_start_queue(dev);
net: use NETDEV_TX_OK instead of 0 in ndo_start_xmit() functions This patch is the result of an automatic spatch transformation to convert all ndo_start_xmit() return values of 0 to NETDEV_TX_OK. Some occurences are missed by the automatic conversion, those will be handled in a seperate patch. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-23 14:03:08 +08:00
return NETDEV_TX_OK;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
length = (length + 1) & ~1;
if (length != skb->len) {
[NET]: Avoid allocating skb in skb_pad First of all it is unnecessary to allocate a new skb in skb_pad since the existing one is not shared. More importantly, our hard_start_xmit interface does not allow a new skb to be allocated since that breaks requeueing. This patch uses pskb_expand_head to expand the existing skb and linearize it if needed. Actually, someone should sift through every instance of skb_pad on a non-linear skb as they do not fit the reasons why this was originally created. Incidentally, this fixes a minor bug when the skb is cloned (tcpdump, TCP, etc.). As it is skb_pad will simply write over a cloned skb. Because of the position of the write it is unlikely to cause problems but still it's best if we don't do it. Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-23 17:06:41 +08:00
if (skb_padto(skb, length))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
goto out;
}
next_ptr = (priv(dev)->tx_head + 1) & 15;
local_irq_save(flags);
if (priv(dev)->tx_tail == next_ptr) {
local_irq_restore(flags);
net: use symbolic values for ndo_start_xmit() return codes Convert magic values 1 and -1 to NETDEV_TX_BUSY and NETDEV_TX_LOCKED respectively. 0 (NETDEV_TX_OK) is not changed to keep the noise down, except in very few cases where its in direct proximity to one of the other values. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-12 14:22:29 +08:00
return NETDEV_TX_BUSY; /* unable to queue */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
ptr = 0x600 * priv(dev)->tx_head;
priv(dev)->tx_head = next_ptr;
next_ptr *= 0x600;
#define TXHDR_FLAGS (TXHDR_TRANSMIT|TXHDR_CHAINCONTINUE|TXHDR_DATAFOLLOWS|TXHDR_ENSUCCESS)
ether3_setbuffer(dev, buffer_write, next_ptr);
ether3_writelong(dev, 0);
ether3_setbuffer(dev, buffer_write, ptr);
ether3_writelong(dev, 0);
ether3_writebuffer(dev, skb->data, length);
ether3_writeword(dev, htons(next_ptr));
ether3_writeword(dev, TXHDR_CHAINCONTINUE >> 16);
ether3_setbuffer(dev, buffer_write, ptr);
ether3_writeword(dev, htons((ptr + length + 4)));
ether3_writeword(dev, TXHDR_FLAGS >> 16);
ether3_ledon(dev);
if (!(ether3_inw(REG_STATUS) & STAT_TXON)) {
ether3_outw(ptr, REG_TRANSMITPTR);
ether3_outw(priv(dev)->regs.command | CMD_TXON, REG_COMMAND);
}
next_ptr = (priv(dev)->tx_head + 1) & 15;
local_irq_restore(flags);
dev_kfree_skb(skb);
if (priv(dev)->tx_tail == next_ptr)
netif_stop_queue(dev);
out:
net: use NETDEV_TX_OK instead of 0 in ndo_start_xmit() functions This patch is the result of an automatic spatch transformation to convert all ndo_start_xmit() return values of 0 to NETDEV_TX_OK. Some occurences are missed by the automatic conversion, those will be handled in a seperate patch. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-23 14:03:08 +08:00
return NETDEV_TX_OK;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
static irqreturn_t
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
ether3_interrupt(int irq, void *dev_id)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct net_device *dev = (struct net_device *)dev_id;
unsigned int status, handled = IRQ_NONE;
#if NET_DEBUG > 1
if(net_debug & DEBUG_INT)
printk("eth3irq: %d ", irq);
#endif
status = ether3_inw(REG_STATUS);
if (status & STAT_INTRX) {
ether3_outw(CMD_ACKINTRX | priv(dev)->regs.command, REG_COMMAND);
ether3_rx(dev, 12);
handled = IRQ_HANDLED;
}
if (status & STAT_INTTX) {
ether3_outw(CMD_ACKINTTX | priv(dev)->regs.command, REG_COMMAND);
ether3_tx(dev);
handled = IRQ_HANDLED;
}
#if NET_DEBUG > 1
if(net_debug & DEBUG_INT)
printk("done\n");
#endif
return handled;
}
/*
* If we have a good packet(s), get it/them out of the buffers.
*/
static int ether3_rx(struct net_device *dev, unsigned int maxcnt)
{
unsigned int next_ptr = priv(dev)->rx_head, received = 0;
ether3_ledon(dev);
do {
unsigned int this_ptr, status;
unsigned char addrs[16];
/*
* read the first 16 bytes from the buffer.
* This contains the status bytes etc and ethernet addresses,
* and we also check the source ethernet address to see if
* it originated from us.
*/
{
unsigned int temp_ptr;
ether3_setbuffer(dev, buffer_read, next_ptr);
temp_ptr = ether3_readword(dev);
status = ether3_readword(dev);
if ((status & (RXSTAT_DONE | RXHDR_CHAINCONTINUE | RXHDR_RECEIVE)) !=
(RXSTAT_DONE | RXHDR_CHAINCONTINUE) || !temp_ptr)
break;
this_ptr = next_ptr + 4;
next_ptr = ntohs(temp_ptr);
}
ether3_setbuffer(dev, buffer_read, this_ptr);
ether3_readbuffer(dev, addrs+2, 12);
if (next_ptr < RX_START || next_ptr >= RX_END) {
int i;
printk("%s: bad next pointer @%04X: ", dev->name, priv(dev)->rx_head);
printk("%02X %02X %02X %02X ", next_ptr >> 8, next_ptr & 255, status & 255, status >> 8);
for (i = 2; i < 14; i++)
printk("%02X ", addrs[i]);
printk("\n");
next_ptr = priv(dev)->rx_head;
break;
}
/*
* ignore our own packets...
*/
if (!(*(unsigned long *)&dev->dev_addr[0] ^ *(unsigned long *)&addrs[2+6]) &&
!(*(unsigned short *)&dev->dev_addr[4] ^ *(unsigned short *)&addrs[2+10])) {
maxcnt ++; /* compensate for loopedback packet */
ether3_outw(next_ptr >> 8, REG_RECVEND);
} else
if (!(status & (RXSTAT_OVERSIZE|RXSTAT_CRCERROR|RXSTAT_DRIBBLEERROR|RXSTAT_SHORTPACKET))) {
unsigned int length = next_ptr - this_ptr;
struct sk_buff *skb;
if (next_ptr <= this_ptr)
length += RX_END - RX_START;
netdev: ethernet dev_alloc_skb to netdev_alloc_skb Replaced deprecating dev_alloc_skb with netdev_alloc_skb in drivers/net/ethernet - Removed extra skb->dev = dev after netdev_alloc_skb Signed-off-by: Pradeep A Dalvi <netdev@pradeepdalvi.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-02-06 19:16:13 +08:00
skb = netdev_alloc_skb(dev, length + 2);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (skb) {
unsigned char *buf;
skb_reserve(skb, 2);
buf = skb_put(skb, length);
ether3_readbuffer(dev, buf + 12, length - 12);
ether3_outw(next_ptr >> 8, REG_RECVEND);
*(unsigned short *)(buf + 0) = *(unsigned short *)(addrs + 2);
*(unsigned long *)(buf + 2) = *(unsigned long *)(addrs + 4);
*(unsigned long *)(buf + 6) = *(unsigned long *)(addrs + 8);
*(unsigned short *)(buf + 10) = *(unsigned short *)(addrs + 12);
skb->protocol = eth_type_trans(skb, dev);
netif_rx(skb);
received ++;
drivers:net: Remove unnecessary OOM messages after netdev_alloc_skb Emitting netdev_alloc_skb and netdev_alloc_skb_ip_align OOM messages is unnecessary as there is already a dump_stack after allocation failures. Other trivial changes around these removals: Convert a few comparisons of pointer to 0 to !pointer. Change flow to remove unnecessary label. Remove now unused variable. Hoist assignment from if. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-08 23:03:25 +08:00
} else {
ether3_outw(next_ptr >> 8, REG_RECVEND);
dev->stats.rx_dropped++;
goto done;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
} else {
ether3: Use net_device_stats from struct net_device struct net_device has its own struct net_device_stats member, so use this one instead of a private copy in the ether1_priv struct. As the new ndo_get_stats function would just return dev->stats we can omit it. This patch also removes an incorrect memset of the stats on open. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 04:15:31 +08:00
struct net_device_stats *stats = &dev->stats;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ether3_outw(next_ptr >> 8, REG_RECVEND);
if (status & RXSTAT_OVERSIZE) stats->rx_over_errors ++;
if (status & RXSTAT_CRCERROR) stats->rx_crc_errors ++;
if (status & RXSTAT_DRIBBLEERROR) stats->rx_fifo_errors ++;
if (status & RXSTAT_SHORTPACKET) stats->rx_length_errors ++;
stats->rx_errors++;
}
}
while (-- maxcnt);
done:
ether3: Use net_device_stats from struct net_device struct net_device has its own struct net_device_stats member, so use this one instead of a private copy in the ether1_priv struct. As the new ndo_get_stats function would just return dev->stats we can omit it. This patch also removes an incorrect memset of the stats on open. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 04:15:31 +08:00
dev->stats.rx_packets += received;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
priv(dev)->rx_head = next_ptr;
/*
* If rx went off line, then that means that the buffer may be full. We
* have dropped at least one packet.
*/
if (!(ether3_inw(REG_STATUS) & STAT_RXON)) {
ether3: Use net_device_stats from struct net_device struct net_device has its own struct net_device_stats member, so use this one instead of a private copy in the ether1_priv struct. As the new ndo_get_stats function would just return dev->stats we can omit it. This patch also removes an incorrect memset of the stats on open. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 04:15:31 +08:00
dev->stats.rx_dropped++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ether3_outw(next_ptr, REG_RECVPTR);
ether3_outw(priv(dev)->regs.command | CMD_RXON, REG_COMMAND);
}
return maxcnt;
}
/*
* Update stats for the transmitted packet(s)
*/
static void ether3_tx(struct net_device *dev)
{
unsigned int tx_tail = priv(dev)->tx_tail;
int max_work = 14;
do {
unsigned long status;
/*
* Read the packet header
*/
ether3_setbuffer(dev, buffer_read, tx_tail * 0x600);
status = ether3_readlong(dev);
/*
* Check to see if this packet has been transmitted
*/
if ((status & (TXSTAT_DONE | TXHDR_TRANSMIT)) !=
(TXSTAT_DONE | TXHDR_TRANSMIT))
break;
/*
* Update errors
*/
if (!(status & (TXSTAT_BABBLED | TXSTAT_16COLLISIONS)))
ether3: Use net_device_stats from struct net_device struct net_device has its own struct net_device_stats member, so use this one instead of a private copy in the ether1_priv struct. As the new ndo_get_stats function would just return dev->stats we can omit it. This patch also removes an incorrect memset of the stats on open. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 04:15:31 +08:00
dev->stats.tx_packets++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
else {
ether3: Use net_device_stats from struct net_device struct net_device has its own struct net_device_stats member, so use this one instead of a private copy in the ether1_priv struct. As the new ndo_get_stats function would just return dev->stats we can omit it. This patch also removes an incorrect memset of the stats on open. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 04:15:31 +08:00
dev->stats.tx_errors++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (status & TXSTAT_16COLLISIONS)
ether3: Use net_device_stats from struct net_device struct net_device has its own struct net_device_stats member, so use this one instead of a private copy in the ether1_priv struct. As the new ndo_get_stats function would just return dev->stats we can omit it. This patch also removes an incorrect memset of the stats on open. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 04:15:31 +08:00
dev->stats.collisions += 16;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (status & TXSTAT_BABBLED)
ether3: Use net_device_stats from struct net_device struct net_device has its own struct net_device_stats member, so use this one instead of a private copy in the ether1_priv struct. As the new ndo_get_stats function would just return dev->stats we can omit it. This patch also removes an incorrect memset of the stats on open. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 04:15:31 +08:00
dev->stats.tx_fifo_errors++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
tx_tail = (tx_tail + 1) & 15;
} while (--max_work);
if (priv(dev)->tx_tail != tx_tail) {
priv(dev)->tx_tail = tx_tail;
netif_wake_queue(dev);
}
}
seeq: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Russell King <linux@arm.linux.org.uk> Cc: linux-arm-kernel@lists.infradead.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 22:23:31 +08:00
static void ether3_banner(void)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
static unsigned version_printed = 0;
if (net_debug && version_printed++ == 0)
printk(KERN_INFO "%s", version);
}
ether3: convert to net_device_ops Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-10 01:29:38 +08:00
static const struct net_device_ops ether3_netdev_ops = {
.ndo_open = ether3_open,
.ndo_stop = ether3_close,
.ndo_start_xmit = ether3_sendpacket,
net: remove use of ndo_set_multicast_list in drivers replace it by ndo_set_rx_mode Signed-off-by: Jiri Pirko <jpirko@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2011-08-16 14:29:01 +08:00
.ndo_set_rx_mode = ether3_setmulticastlist,
ether3: convert to net_device_ops Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-10 01:29:38 +08:00
.ndo_tx_timeout = ether3_timeout,
.ndo_validate_addr = eth_validate_addr,
.ndo_change_mtu = eth_change_mtu,
.ndo_set_mac_address = eth_mac_addr,
};
seeq: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Russell King <linux@arm.linux.org.uk> Cc: linux-arm-kernel@lists.infradead.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 22:23:31 +08:00
static int
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ether3_probe(struct expansion_card *ec, const struct ecard_id *id)
{
const struct ether3_data *data = id->data;
struct net_device *dev;
[NET]: Introduce and use print_mac() and DECLARE_MAC_BUF() This is nicer than the MAC_FMT stuff. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:59:30 +08:00
int bus_type, ret;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ether3_banner();
ret = ecard_request_resources(ec);
if (ret)
goto out;
dev = alloc_etherdev(sizeof(struct dev_priv));
if (!dev) {
ret = -ENOMEM;
goto release;
}
SET_NETDEV_DEV(dev, &ec->dev);
[ARM] ecard: add ecardm_iomap() / ecardm_iounmap() Add devres ecardm_iomap() and ecardm_iounmap() for Acorn expansion cards. Convert all expansion card drivers to use them. Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2007-05-11 01:40:51 +08:00
priv(dev)->base = ecardm_iomap(ec, ECARD_RES_MEMC, 0, 0);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (!priv(dev)->base) {
ret = -ENOMEM;
goto free;
}
ec->irqaddr = priv(dev)->base + data->base_offset;
ec->irqmask = 0xf0;
priv(dev)->seeq = priv(dev)->base + data->base_offset;
dev->irq = ec->irq;
ether3_addr(dev->dev_addr, ec);
init_timer(&priv(dev)->timer);
/* Reset card...
*/
ether3_outb(0x80, REG_CONFIG2 + 4);
bus_type = BUS_UNKNOWN;
udelay(4);
/* Test using Receive Pointer (16-bit register) to find out
* how the ether3 is connected to the bus...
*/
if (ether3_probe_bus_8(dev, 0x100) &&
ether3_probe_bus_8(dev, 0x201))
bus_type = BUS_8;
if (bus_type == BUS_UNKNOWN &&
ether3_probe_bus_16(dev, 0x101) &&
ether3_probe_bus_16(dev, 0x201))
bus_type = BUS_16;
switch (bus_type) {
case BUS_UNKNOWN:
printk(KERN_ERR "%s: unable to identify bus width\n", dev->name);
ret = -ENODEV;
goto free;
case BUS_8:
printk(KERN_ERR "%s: %s found, but is an unsupported "
"8-bit card\n", dev->name, data->name);
ret = -ENODEV;
goto free;
default:
break;
}
if (ether3_init_2(dev)) {
ret = -ENODEV;
goto free;
}
ether3: convert to net_device_ops Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-10 01:29:38 +08:00
dev->netdev_ops = &ether3_netdev_ops;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
dev->watchdog_timeo = 5 * HZ / 100;
ret = register_netdev(dev);
if (ret)
goto free;
net: convert print_mac to %pM This converts pretty much everything to print_mac. There were a few things that had conflicts which I have just dropped for now, no harm done. I've built an allyesconfig with this and looked at the files that weren't built very carefully, but it's a huge patch. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-28 06:59:26 +08:00
printk("%s: %s in slot %d, %pM\n",
dev->name, data->name, ec->slot_no, dev->dev_addr);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ecard_set_drvdata(ec, dev);
return 0;
free:
free_netdev(dev);
release:
ecard_release_resources(ec);
out:
return ret;
}
seeq: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Russell King <linux@arm.linux.org.uk> Cc: linux-arm-kernel@lists.infradead.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 22:23:31 +08:00
static void ether3_remove(struct expansion_card *ec)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct net_device *dev = ecard_get_drvdata(ec);
ecard_set_drvdata(ec, NULL);
unregister_netdev(dev);
free_netdev(dev);
ecard_release_resources(ec);
}
static struct ether3_data ether3 = {
.name = "ether3",
.base_offset = 0,
};
static struct ether3_data etherb = {
.name = "etherb",
.base_offset = 0x800,
};
static const struct ecard_id ether3_ids[] = {
{ MANU_ANT2, PROD_ANT_ETHER3, &ether3 },
{ MANU_ANT, PROD_ANT_ETHER3, &ether3 },
{ MANU_ANT, PROD_ANT_ETHERB, &etherb },
{ 0xffff, 0xffff }
};
static struct ecard_driver ether3_driver = {
.probe = ether3_probe,
seeq: remove __dev* attributes CONFIG_HOTPLUG is going away as an option. As result the __dev* markings will be going away. Remove use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit. Signed-off-by: Bill Pemberton <wfp5p@virginia.edu> Cc: Russell King <linux@arm.linux.org.uk> Cc: linux-arm-kernel@lists.infradead.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-03 22:23:31 +08:00
.remove = ether3_remove,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
.id_table = ether3_ids,
.drv = {
.name = "ether3",
},
};
static int __init ether3_init(void)
{
return ecard_register_driver(&ether3_driver);
}
static void __exit ether3_exit(void)
{
ecard_remove_driver(&ether3_driver);
}
module_init(ether3_init);
module_exit(ether3_exit);
MODULE_LICENSE("GPL");