linux_dsm_epyc7002/drivers/net/wan/hdlc_cisco.c
Krzysztof Halasa 9652041da1 WAN: fix Cisco HDLC handshaking.
Cisco HDLC uses keepalive packets and sequence numbers to determine link
state. In rare cases both ends could transmit keepalive packets at the same
time, causing the received sequence numbers to be treated as incorrect.
Now we accept our current sequence number as well as the previous one.

Signed-off-by: Krzysztof Hałasa <khc@pm.waw.pl>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 03:16:53 -07:00

413 lines
9.7 KiB
C

/*
* Generic HDLC support routines for Linux
* Cisco HDLC support
*
* Copyright (C) 2000 - 2006 Krzysztof Halasa <khc@pm.waw.pl>
*
* This program is free software; you can redistribute it and/or modify it
* under the terms of version 2 of the GNU General Public License
* as published by the Free Software Foundation.
*/
#include <linux/errno.h>
#include <linux/hdlc.h>
#include <linux/if_arp.h>
#include <linux/inetdevice.h>
#include <linux/init.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/pkt_sched.h>
#include <linux/poll.h>
#include <linux/rtnetlink.h>
#include <linux/skbuff.h>
#include <linux/slab.h>
#undef DEBUG_HARD_HEADER
#define CISCO_MULTICAST 0x8F /* Cisco multicast address */
#define CISCO_UNICAST 0x0F /* Cisco unicast address */
#define CISCO_KEEPALIVE 0x8035 /* Cisco keepalive protocol */
#define CISCO_SYS_INFO 0x2000 /* Cisco interface/system info */
#define CISCO_ADDR_REQ 0 /* Cisco address request */
#define CISCO_ADDR_REPLY 1 /* Cisco address reply */
#define CISCO_KEEPALIVE_REQ 2 /* Cisco keepalive request */
struct hdlc_header {
u8 address;
u8 control;
__be16 protocol;
}__attribute__ ((packed));
struct cisco_packet {
__be32 type; /* code */
__be32 par1;
__be32 par2;
__be16 rel; /* reliability */
__be32 time;
}__attribute__ ((packed));
#define CISCO_PACKET_LEN 18
#define CISCO_BIG_PACKET_LEN 20
struct cisco_state {
cisco_proto settings;
struct timer_list timer;
spinlock_t lock;
unsigned long last_poll;
int up;
u32 txseq; /* TX sequence number, 0 = none */
u32 rxseq; /* RX sequence number */
};
static int cisco_ioctl(struct net_device *dev, struct ifreq *ifr);
static inline struct cisco_state* state(hdlc_device *hdlc)
{
return (struct cisco_state *)hdlc->state;
}
static int cisco_hard_header(struct sk_buff *skb, struct net_device *dev,
u16 type, const void *daddr, const void *saddr,
unsigned int len)
{
struct hdlc_header *data;
#ifdef DEBUG_HARD_HEADER
printk(KERN_DEBUG "%s: cisco_hard_header called\n", dev->name);
#endif
skb_push(skb, sizeof(struct hdlc_header));
data = (struct hdlc_header*)skb->data;
if (type == CISCO_KEEPALIVE)
data->address = CISCO_MULTICAST;
else
data->address = CISCO_UNICAST;
data->control = 0;
data->protocol = htons(type);
return sizeof(struct hdlc_header);
}
static void cisco_keepalive_send(struct net_device *dev, u32 type,
__be32 par1, __be32 par2)
{
struct sk_buff *skb;
struct cisco_packet *data;
skb = dev_alloc_skb(sizeof(struct hdlc_header) +
sizeof(struct cisco_packet));
if (!skb) {
printk(KERN_WARNING
"%s: Memory squeeze on cisco_keepalive_send()\n",
dev->name);
return;
}
skb_reserve(skb, 4);
cisco_hard_header(skb, dev, CISCO_KEEPALIVE, NULL, NULL, 0);
data = (struct cisco_packet*)(skb->data + 4);
data->type = htonl(type);
data->par1 = par1;
data->par2 = par2;
data->rel = cpu_to_be16(0xFFFF);
/* we will need do_div here if 1000 % HZ != 0 */
data->time = htonl((jiffies - INITIAL_JIFFIES) * (1000 / HZ));
skb_put(skb, sizeof(struct cisco_packet));
skb->priority = TC_PRIO_CONTROL;
skb->dev = dev;
skb_reset_network_header(skb);
dev_queue_xmit(skb);
}
static __be16 cisco_type_trans(struct sk_buff *skb, struct net_device *dev)
{
struct hdlc_header *data = (struct hdlc_header*)skb->data;
if (skb->len < sizeof(struct hdlc_header))
return cpu_to_be16(ETH_P_HDLC);
if (data->address != CISCO_MULTICAST &&
data->address != CISCO_UNICAST)
return cpu_to_be16(ETH_P_HDLC);
switch(data->protocol) {
case cpu_to_be16(ETH_P_IP):
case cpu_to_be16(ETH_P_IPX):
case cpu_to_be16(ETH_P_IPV6):
skb_pull(skb, sizeof(struct hdlc_header));
return data->protocol;
default:
return cpu_to_be16(ETH_P_HDLC);
}
}
static int cisco_rx(struct sk_buff *skb)
{
struct net_device *dev = skb->dev;
hdlc_device *hdlc = dev_to_hdlc(dev);
struct cisco_state *st = state(hdlc);
struct hdlc_header *data = (struct hdlc_header*)skb->data;
struct cisco_packet *cisco_data;
struct in_device *in_dev;
__be32 addr, mask;
u32 ack;
if (skb->len < sizeof(struct hdlc_header))
goto rx_error;
if (data->address != CISCO_MULTICAST &&
data->address != CISCO_UNICAST)
goto rx_error;
switch (ntohs(data->protocol)) {
case CISCO_SYS_INFO:
/* Packet is not needed, drop it. */
dev_kfree_skb_any(skb);
return NET_RX_SUCCESS;
case CISCO_KEEPALIVE:
if ((skb->len != sizeof(struct hdlc_header) +
CISCO_PACKET_LEN) &&
(skb->len != sizeof(struct hdlc_header) +
CISCO_BIG_PACKET_LEN)) {
printk(KERN_INFO "%s: Invalid length of Cisco control"
" packet (%d bytes)\n", dev->name, skb->len);
goto rx_error;
}
cisco_data = (struct cisco_packet*)(skb->data + sizeof
(struct hdlc_header));
switch(ntohl (cisco_data->type)) {
case CISCO_ADDR_REQ: /* Stolen from syncppp.c :-) */
in_dev = dev->ip_ptr;
addr = 0;
mask = ~cpu_to_be32(0); /* is the mask correct? */
if (in_dev != NULL) {
struct in_ifaddr **ifap = &in_dev->ifa_list;
while (*ifap != NULL) {
if (strcmp(dev->name,
(*ifap)->ifa_label) == 0) {
addr = (*ifap)->ifa_local;
mask = (*ifap)->ifa_mask;
break;
}
ifap = &(*ifap)->ifa_next;
}
cisco_keepalive_send(dev, CISCO_ADDR_REPLY,
addr, mask);
}
dev_kfree_skb_any(skb);
return NET_RX_SUCCESS;
case CISCO_ADDR_REPLY:
printk(KERN_INFO "%s: Unexpected Cisco IP address "
"reply\n", dev->name);
goto rx_error;
case CISCO_KEEPALIVE_REQ:
spin_lock(&st->lock);
st->rxseq = ntohl(cisco_data->par1);
ack = ntohl(cisco_data->par2);
if (ack && (ack == st->txseq ||
/* our current REQ may be in transit */
ack == st->txseq - 1)) {
st->last_poll = jiffies;
if (!st->up) {
u32 sec, min, hrs, days;
sec = ntohl(cisco_data->time) / 1000;
min = sec / 60; sec -= min * 60;
hrs = min / 60; min -= hrs * 60;
days = hrs / 24; hrs -= days * 24;
printk(KERN_INFO "%s: Link up (peer "
"uptime %ud%uh%um%us)\n",
dev->name, days, hrs, min, sec);
netif_dormant_off(dev);
st->up = 1;
}
}
spin_unlock(&st->lock);
dev_kfree_skb_any(skb);
return NET_RX_SUCCESS;
} /* switch(keepalive type) */
} /* switch(protocol) */
printk(KERN_INFO "%s: Unsupported protocol %x\n", dev->name,
ntohs(data->protocol));
dev_kfree_skb_any(skb);
return NET_RX_DROP;
rx_error:
dev->stats.rx_errors++; /* Mark error */
dev_kfree_skb_any(skb);
return NET_RX_DROP;
}
static void cisco_timer(unsigned long arg)
{
struct net_device *dev = (struct net_device *)arg;
hdlc_device *hdlc = dev_to_hdlc(dev);
struct cisco_state *st = state(hdlc);
spin_lock(&st->lock);
if (st->up &&
time_after(jiffies, st->last_poll + st->settings.timeout * HZ)) {
st->up = 0;
printk(KERN_INFO "%s: Link down\n", dev->name);
netif_dormant_on(dev);
}
cisco_keepalive_send(dev, CISCO_KEEPALIVE_REQ, htonl(++st->txseq),
htonl(st->rxseq));
spin_unlock(&st->lock);
st->timer.expires = jiffies + st->settings.interval * HZ;
st->timer.function = cisco_timer;
st->timer.data = arg;
add_timer(&st->timer);
}
static void cisco_start(struct net_device *dev)
{
hdlc_device *hdlc = dev_to_hdlc(dev);
struct cisco_state *st = state(hdlc);
unsigned long flags;
spin_lock_irqsave(&st->lock, flags);
st->up = st->txseq = st->rxseq = 0;
spin_unlock_irqrestore(&st->lock, flags);
init_timer(&st->timer);
st->timer.expires = jiffies + HZ; /* First poll after 1 s */
st->timer.function = cisco_timer;
st->timer.data = (unsigned long)dev;
add_timer(&st->timer);
}
static void cisco_stop(struct net_device *dev)
{
hdlc_device *hdlc = dev_to_hdlc(dev);
struct cisco_state *st = state(hdlc);
unsigned long flags;
del_timer_sync(&st->timer);
spin_lock_irqsave(&st->lock, flags);
netif_dormant_on(dev);
st->up = st->txseq = 0;
spin_unlock_irqrestore(&st->lock, flags);
}
static struct hdlc_proto proto = {
.start = cisco_start,
.stop = cisco_stop,
.type_trans = cisco_type_trans,
.ioctl = cisco_ioctl,
.netif_rx = cisco_rx,
.module = THIS_MODULE,
};
static const struct header_ops cisco_header_ops = {
.create = cisco_hard_header,
};
static int cisco_ioctl(struct net_device *dev, struct ifreq *ifr)
{
cisco_proto __user *cisco_s = ifr->ifr_settings.ifs_ifsu.cisco;
const size_t size = sizeof(cisco_proto);
cisco_proto new_settings;
hdlc_device *hdlc = dev_to_hdlc(dev);
int result;
switch (ifr->ifr_settings.type) {
case IF_GET_PROTO:
if (dev_to_hdlc(dev)->proto != &proto)
return -EINVAL;
ifr->ifr_settings.type = IF_PROTO_CISCO;
if (ifr->ifr_settings.size < size) {
ifr->ifr_settings.size = size; /* data size wanted */
return -ENOBUFS;
}
if (copy_to_user(cisco_s, &state(hdlc)->settings, size))
return -EFAULT;
return 0;
case IF_PROTO_CISCO:
if (!capable(CAP_NET_ADMIN))
return -EPERM;
if (dev->flags & IFF_UP)
return -EBUSY;
if (copy_from_user(&new_settings, cisco_s, size))
return -EFAULT;
if (new_settings.interval < 1 ||
new_settings.timeout < 2)
return -EINVAL;
result = hdlc->attach(dev, ENCODING_NRZ,PARITY_CRC16_PR1_CCITT);
if (result)
return result;
result = attach_hdlc_protocol(dev, &proto,
sizeof(struct cisco_state));
if (result)
return result;
memcpy(&state(hdlc)->settings, &new_settings, size);
spin_lock_init(&state(hdlc)->lock);
dev->header_ops = &cisco_header_ops;
dev->type = ARPHRD_CISCO;
netif_dormant_on(dev);
return 0;
}
return -EINVAL;
}
static int __init mod_init(void)
{
register_hdlc_protocol(&proto);
return 0;
}
static void __exit mod_exit(void)
{
unregister_hdlc_protocol(&proto);
}
module_init(mod_init);
module_exit(mod_exit);
MODULE_AUTHOR("Krzysztof Halasa <khc@pm.waw.pl>");
MODULE_DESCRIPTION("Cisco HDLC protocol support for generic HDLC");
MODULE_LICENSE("GPL v2");