WAN: syncppp.c is no longer used by any kernel code. Remove it.
authorKrzysztof Hałasa <khc@pm.waw.pl>
Thu, 14 Aug 2008 17:18:17 +0000 (19:18 +0200)
committerKrzysztof Hałasa <khc@pm.waw.pl>
Sat, 22 Nov 2008 01:49:48 +0000 (02:49 +0100)
Signed-off-by: Krzysztof Hałasa <khc@pm.waw.pl>
Documentation/DocBook/Makefile
Documentation/DocBook/networking.tmpl
Documentation/DocBook/wanbook.tmpl [deleted file]
drivers/net/wan/syncppp.c [deleted file]
include/net/syncppp.h [deleted file]

index 9b1f6ca..0a08126 100644 (file)
@@ -6,7 +6,7 @@
 # To add a new book the only step required is to add the book to the
 # list of DOCBOOKS.
 
-DOCBOOKS := wanbook.xml z8530book.xml mcabook.xml \
+DOCBOOKS := z8530book.xml mcabook.xml \
            kernel-hacking.xml kernel-locking.xml deviceiobook.xml \
            procfs-guide.xml writing_usb_driver.xml networking.xml \
            kernel-api.xml filesystems.xml lsm.xml usb.xml kgdb.xml \
index f24f9e8..627707a 100644 (file)
@@ -98,9 +98,6 @@
 X!Enet/core/wireless.c
      </sect1>
 -->
-     <sect1><title>Synchronous PPP</title>
-!Edrivers/net/wan/syncppp.c
-     </sect1>
   </chapter>
 
 </book>
diff --git a/Documentation/DocBook/wanbook.tmpl b/Documentation/DocBook/wanbook.tmpl
deleted file mode 100644 (file)
index 8c93db1..0000000
+++ /dev/null
@@ -1,99 +0,0 @@
-<?xml version="1.0" encoding="UTF-8"?>
-<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN"
-       "http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd" []>
-
-<book id="WANGuide">
- <bookinfo>
-  <title>Synchronous PPP and Cisco HDLC Programming Guide</title>
-  
-  <authorgroup>
-   <author>
-    <firstname>Alan</firstname>
-    <surname>Cox</surname>
-    <affiliation>
-     <address>
-      <email>alan@lxorguk.ukuu.org.uk</email>
-     </address>
-    </affiliation>
-   </author>
-  </authorgroup>
-
-  <copyright>
-   <year>2000</year>
-   <holder>Alan Cox</holder>
-  </copyright>
-
-  <legalnotice>
-   <para>
-     This documentation is free software; you can redistribute
-     it and/or modify it under the terms of the GNU General Public
-     License as published by the Free Software Foundation; either
-     version 2 of the License, or (at your option) any later
-     version.
-   </para>
-      
-   <para>
-     This program is distributed in the hope that it will be
-     useful, but WITHOUT ANY WARRANTY; without even the implied
-     warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
-     See the GNU General Public License for more details.
-   </para>
-      
-   <para>
-     You should have received a copy of the GNU General Public
-     License along with this program; if not, write to the Free
-     Software Foundation, Inc., 59 Temple Place, Suite 330, Boston,
-     MA 02111-1307 USA
-   </para>
-      
-   <para>
-     For more details see the file COPYING in the source
-     distribution of Linux.
-   </para>
-  </legalnotice>
- </bookinfo>
-
-<toc></toc>
-
-  <chapter id="intro">
-      <title>Introduction</title>
-  <para>
-       The syncppp drivers in Linux provide a fairly complete 
-       implementation of Cisco HDLC and a minimal implementation of
-       PPP. The longer term goal is to switch the PPP layer to the
-       generic PPP interface that is new in Linux 2.3.x. The API should
-       remain unchanged when this is done, but support will then be
-       available for IPX, compression and other PPP features
-  </para>
-  </chapter>
-  <chapter id="bugs">
-     <title>Known Bugs And Assumptions</title>
-  <para>
-  <variablelist>
-    <varlistentry><term>PPP is minimal</term>
-    <listitem>
-    <para>
-       The current PPP implementation is very basic, although sufficient
-       for most wan usages.
-    </para>
-    </listitem></varlistentry>
-
-    <varlistentry><term>Cisco HDLC Quirks</term>
-    <listitem>
-    <para>
-       Currently we do not end all packets with the correct Cisco multicast
-       or unicast flags. Nothing appears to mind too much but this should
-       be corrected.
-    </para>
-    </listitem></varlistentry>
-  </variablelist>
-       
-  </para>
-  </chapter>
-
-  <chapter id="pubfunctions">
-     <title>Public Functions Provided</title>
-!Edrivers/net/wan/syncppp.c
-  </chapter>
-
-</book>
diff --git a/drivers/net/wan/syncppp.c b/drivers/net/wan/syncppp.c
deleted file mode 100644 (file)
index 58ae8a2..0000000
+++ /dev/null
@@ -1,1476 +0,0 @@
-/*
- *     NET3:   A (fairly minimal) implementation of synchronous PPP for Linux
- *             as well as a CISCO HDLC implementation. See the copyright 
- *             message below for the original source.
- *
- *     This program is free software; you can redistribute it and/or
- *     modify it under the terms of the GNU General Public License
- *     as published by the Free Software Foundation; either version
- *     2 of the license, or (at your option) any later version.
- *
- *     Note however. This code is also used in a different form by FreeBSD.
- *     Therefore when making any non OS specific change please consider
- *     contributing it back to the original author under the terms
- *     below in addition.
- *             -- Alan
- *
- *     Port for Linux-2.1 by Jan "Yenya" Kasprzak <kas@fi.muni.cz>
- */
-
-/*
- * Synchronous PPP/Cisco link level subroutines.
- * Keepalive protocol implemented in both Cisco and PPP modes.
- *
- * Copyright (C) 1994 Cronyx Ltd.
- * Author: Serge Vakulenko, <vak@zebub.msk.su>
- *
- * This software is distributed with NO WARRANTIES, not even the implied
- * warranties for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
- *
- * Authors grant any other persons or organisations permission to use
- * or modify this software as long as this message is kept with the software,
- * all derivative works or modified versions.
- *
- * Version 1.9, Wed Oct  4 18:58:15 MSK 1995
- *
- * $Id: syncppp.c,v 1.18 2000/04/11 05:25:31 asj Exp $
- */
-#undef DEBUG
-
-#include <linux/module.h>
-#include <linux/kernel.h>
-#include <linux/errno.h>
-#include <linux/init.h>
-#include <linux/if_arp.h>
-#include <linux/skbuff.h>
-#include <linux/route.h>
-#include <linux/netdevice.h>
-#include <linux/inetdevice.h>
-#include <linux/random.h>
-#include <linux/pkt_sched.h>
-#include <linux/spinlock.h>
-#include <linux/rcupdate.h>
-
-#include <net/net_namespace.h>
-#include <net/syncppp.h>
-
-#include <asm/byteorder.h>
-#include <asm/uaccess.h>
-
-#define MAXALIVECNT     6               /* max. alive packets */
-
-#define PPP_ALLSTATIONS 0xff            /* All-Stations broadcast address */
-#define PPP_UI          0x03            /* Unnumbered Information */
-#define PPP_IP          0x0021          /* Internet Protocol */
-#define PPP_ISO         0x0023          /* ISO OSI Protocol */
-#define PPP_XNS         0x0025          /* Xerox NS Protocol */
-#define PPP_IPX         0x002b          /* Novell IPX Protocol */
-#define PPP_LCP         0xc021          /* Link Control Protocol */
-#define PPP_IPCP        0x8021          /* Internet Protocol Control Protocol */
-
-#define LCP_CONF_REQ    1               /* PPP LCP configure request */
-#define LCP_CONF_ACK    2               /* PPP LCP configure acknowledge */
-#define LCP_CONF_NAK    3               /* PPP LCP configure negative ack */
-#define LCP_CONF_REJ    4               /* PPP LCP configure reject */
-#define LCP_TERM_REQ    5               /* PPP LCP terminate request */
-#define LCP_TERM_ACK    6               /* PPP LCP terminate acknowledge */
-#define LCP_CODE_REJ    7               /* PPP LCP code reject */
-#define LCP_PROTO_REJ   8               /* PPP LCP protocol reject */
-#define LCP_ECHO_REQ    9               /* PPP LCP echo request */
-#define LCP_ECHO_REPLY  10              /* PPP LCP echo reply */
-#define LCP_DISC_REQ    11              /* PPP LCP discard request */
-
-#define LCP_OPT_MRU             1       /* maximum receive unit */
-#define LCP_OPT_ASYNC_MAP       2       /* async control character map */
-#define LCP_OPT_AUTH_PROTO      3       /* authentication protocol */
-#define LCP_OPT_QUAL_PROTO      4       /* quality protocol */
-#define LCP_OPT_MAGIC           5       /* magic number */
-#define LCP_OPT_RESERVED        6       /* reserved */
-#define LCP_OPT_PROTO_COMP      7       /* protocol field compression */
-#define LCP_OPT_ADDR_COMP       8       /* address/control field compression */
-
-#define IPCP_CONF_REQ   LCP_CONF_REQ    /* PPP IPCP configure request */
-#define IPCP_CONF_ACK   LCP_CONF_ACK    /* PPP IPCP configure acknowledge */
-#define IPCP_CONF_NAK   LCP_CONF_NAK    /* PPP IPCP configure negative ack */
-#define IPCP_CONF_REJ   LCP_CONF_REJ    /* PPP IPCP configure reject */
-#define IPCP_TERM_REQ   LCP_TERM_REQ    /* PPP IPCP terminate request */
-#define IPCP_TERM_ACK   LCP_TERM_ACK    /* PPP IPCP terminate acknowledge */
-#define IPCP_CODE_REJ   LCP_CODE_REJ    /* PPP IPCP code reject */
-
-#define CISCO_MULTICAST         0x8f    /* Cisco multicast address */
-#define CISCO_UNICAST           0x0f    /* Cisco unicast address */
-#define CISCO_KEEPALIVE         0x8035  /* Cisco keepalive protocol */
-#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 ppp_header {
-       u8 address;
-       u8 control;
-       __be16 protocol;
-};
-#define PPP_HEADER_LEN          sizeof (struct ppp_header)
-
-struct lcp_header {
-       u8 type;
-       u8 ident;
-       __be16 len;
-};
-#define LCP_HEADER_LEN          sizeof (struct lcp_header)
-
-struct cisco_packet {
-       __be32 type;
-       __be32 par1;
-       __be32 par2;
-       __be16 rel;
-       __be16 time0;
-       __be16 time1;
-};
-#define CISCO_PACKET_LEN 18
-#define CISCO_BIG_PACKET_LEN 20
-
-static struct sppp *spppq;
-static struct timer_list sppp_keepalive_timer;
-static DEFINE_SPINLOCK(spppq_lock);
-
-/* global xmit queue for sending packets while spinlock is held */
-static struct sk_buff_head tx_queue;
-
-static void sppp_keepalive (unsigned long dummy);
-static void sppp_cp_send (struct sppp *sp, u16 proto, u8 type,
-       u8 ident, u16 len, void *data);
-static void sppp_cisco_send (struct sppp *sp, int type, u32 par1, u32 par2);
-static void sppp_lcp_input (struct sppp *sp, struct sk_buff *m);
-static void sppp_cisco_input (struct sppp *sp, struct sk_buff *m);
-static void sppp_ipcp_input (struct sppp *sp, struct sk_buff *m);
-static void sppp_lcp_open (struct sppp *sp);
-static void sppp_ipcp_open (struct sppp *sp);
-static int sppp_lcp_conf_parse_options (struct sppp *sp, struct lcp_header *h,
-       int len, u32 *magic);
-static void sppp_cp_timeout (unsigned long arg);
-static char *sppp_lcp_type_name (u8 type);
-static char *sppp_ipcp_type_name (u8 type);
-static void sppp_print_bytes (u8 *p, u16 len);
-
-static int debug;
-
-/* Flush global outgoing packet queue to dev_queue_xmit().
- *
- * dev_queue_xmit() must be called with interrupts enabled
- * which means it can't be called with spinlocks held.
- * If a packet needs to be sent while a spinlock is held,
- * then put the packet into tx_queue, and call sppp_flush_xmit()
- * after spinlock is released.
- */
-static void sppp_flush_xmit(void)
-{
-       struct sk_buff *skb;
-       while ((skb = skb_dequeue(&tx_queue)) != NULL)
-               dev_queue_xmit(skb);
-}
-
-/*
- *     Interface down stub
- */    
-
-static void if_down(struct net_device *dev)
-{
-       struct sppp *sp = (struct sppp *)sppp_of(dev);
-
-       sp->pp_link_state=SPPP_LINK_DOWN;
-}
-
-/*
- * Timeout routine activations.
- */
-
-static void sppp_set_timeout(struct sppp *p,int s) 
-{
-       if (! (p->pp_flags & PP_TIMO)) 
-       {
-               init_timer(&p->pp_timer);
-               p->pp_timer.function=sppp_cp_timeout;
-               p->pp_timer.expires=jiffies+s*HZ;
-               p->pp_timer.data=(unsigned long)p;
-               p->pp_flags |= PP_TIMO;
-               add_timer(&p->pp_timer);
-       }
-}
-
-static void sppp_clear_timeout(struct sppp *p)
-{
-       if (p->pp_flags & PP_TIMO) 
-       {
-               del_timer(&p->pp_timer);
-               p->pp_flags &= ~PP_TIMO; 
-       }
-}
-
-/**
- *     sppp_input -    receive and process a WAN PPP frame
- *     @skb:   The buffer to process
- *     @dev:   The device it arrived on
- *
- *     This can be called directly by cards that do not have
- *     timing constraints but is normally called from the network layer
- *     after interrupt servicing to process frames queued via netif_rx().
- *
- *     We process the options in the card. If the frame is destined for
- *     the protocol stacks then it requeues the frame for the upper level
- *     protocol. If it is a control from it is processed and discarded
- *     here.
- */
-static void sppp_input (struct net_device *dev, struct sk_buff *skb)
-{
-       struct ppp_header *h;
-       struct sppp *sp = (struct sppp *)sppp_of(dev);
-       unsigned long flags;
-
-       skb->dev=dev;
-       skb_reset_mac_header(skb);
-
-       if (!pskb_may_pull(skb, PPP_HEADER_LEN)) {
-               /* Too small packet, drop it. */
-               if (sp->pp_flags & PP_DEBUG)
-                       printk (KERN_DEBUG "%s: input packet is too small, %d bytes\n",
-                               dev->name, skb->len);
-               kfree_skb(skb);
-               return;
-       }
-
-       /* Get PPP header. */
-       h = (struct ppp_header *)skb->data;
-       skb_pull(skb,sizeof(struct ppp_header));
-
-       spin_lock_irqsave(&sp->lock, flags);
-       
-       switch (h->address) {
-       default:        /* Invalid PPP packet. */
-               goto invalid;
-       case PPP_ALLSTATIONS:
-               if (h->control != PPP_UI)
-                       goto invalid;
-               if (sp->pp_flags & PP_CISCO) {
-                       if (sp->pp_flags & PP_DEBUG)
-                               printk (KERN_WARNING "%s: PPP packet in Cisco mode <0x%x 0x%x 0x%x>\n",
-                                       dev->name,
-                                       h->address, h->control, ntohs (h->protocol));
-                       goto drop;
-               }
-               switch (ntohs (h->protocol)) {
-               default:
-                       if (sp->lcp.state == LCP_STATE_OPENED)
-                               sppp_cp_send (sp, PPP_LCP, LCP_PROTO_REJ,
-                                       ++sp->pp_seq, skb->len + 2,
-                                       &h->protocol);
-                       if (sp->pp_flags & PP_DEBUG)
-                               printk (KERN_WARNING "%s: invalid input protocol <0x%x 0x%x 0x%x>\n",
-                                       dev->name,
-                                       h->address, h->control, ntohs (h->protocol));
-                       goto drop;
-               case PPP_LCP:
-                       sppp_lcp_input (sp, skb);
-                       goto drop;
-               case PPP_IPCP:
-                       if (sp->lcp.state == LCP_STATE_OPENED)
-                               sppp_ipcp_input (sp, skb);
-                       else
-                               printk(KERN_DEBUG "IPCP when still waiting LCP finish.\n");
-                       goto drop;
-               case PPP_IP:
-                       if (sp->ipcp.state == IPCP_STATE_OPENED) {
-                               if(sp->pp_flags&PP_DEBUG)
-                                       printk(KERN_DEBUG "Yow an IP frame.\n");
-                               skb->protocol=htons(ETH_P_IP);
-                               netif_rx(skb);
-                               goto done;
-                       }
-                       break;
-#ifdef IPX
-               case PPP_IPX:
-                       /* IPX IPXCP not implemented yet */
-                       if (sp->lcp.state == LCP_STATE_OPENED) {
-                               skb->protocol=htons(ETH_P_IPX);
-                               netif_rx(skb);
-                               goto done;
-                       }
-                       break;
-#endif
-               }
-               break;
-       case CISCO_MULTICAST:
-       case CISCO_UNICAST:
-               /* Don't check the control field here (RFC 1547). */
-               if (! (sp->pp_flags & PP_CISCO)) {
-                       if (sp->pp_flags & PP_DEBUG)
-                               printk (KERN_WARNING "%s: Cisco packet in PPP mode <0x%x 0x%x 0x%x>\n",
-                                       dev->name,
-                                       h->address, h->control, ntohs (h->protocol));
-                       goto drop;
-               }
-               switch (ntohs (h->protocol)) {
-               default:
-                       goto invalid;
-               case CISCO_KEEPALIVE:
-                       sppp_cisco_input (sp, skb);
-                       goto drop;
-#ifdef CONFIG_INET
-               case ETH_P_IP:
-                       skb->protocol=htons(ETH_P_IP);
-                       netif_rx(skb);
-                       goto done;
-#endif
-#ifdef CONFIG_IPX
-               case ETH_P_IPX:
-                       skb->protocol=htons(ETH_P_IPX);
-                       netif_rx(skb);
-                       goto done;
-#endif
-               }
-               break;
-       }
-       goto drop;
-
-invalid:
-       if (sp->pp_flags & PP_DEBUG)
-               printk (KERN_WARNING "%s: invalid input packet <0x%x 0x%x 0x%x>\n",
-                       dev->name, h->address, h->control, ntohs (h->protocol));
-drop:
-       kfree_skb(skb);
-done:
-       spin_unlock_irqrestore(&sp->lock, flags);
-       sppp_flush_xmit();
-       return;
-}
-
-/*
- *     Handle transmit packets.
- */
-static int sppp_hard_header(struct sk_buff *skb,
-                           struct net_device *dev, __u16 type,
-                           const void *daddr, const void *saddr,
-                           unsigned int len)
-{
-       struct sppp *sp = (struct sppp *)sppp_of(dev);
-       struct ppp_header *h;
-       skb_push(skb,sizeof(struct ppp_header));
-       h=(struct ppp_header *)skb->data;
-       if(sp->pp_flags&PP_CISCO)
-       {
-               h->address = CISCO_UNICAST;
-               h->control = 0;
-       }
-       else
-       {
-               h->address = PPP_ALLSTATIONS;
-               h->control = PPP_UI;
-       }
-       if(sp->pp_flags & PP_CISCO)
-       {
-               h->protocol = htons(type);
-       }
-       else switch(type)
-       {
-               case ETH_P_IP:
-                       h->protocol = htons(PPP_IP);
-                       break;
-               case ETH_P_IPX:
-                       h->protocol = htons(PPP_IPX);
-                       break;
-       }
-       return sizeof(struct ppp_header);
-}
-
-static const struct header_ops sppp_header_ops = {
-       .create = sppp_hard_header,
-};
-
-/*
- * Send keepalive packets, every 10 seconds.
- */
-
-static void sppp_keepalive (unsigned long dummy)
-{
-       struct sppp *sp;
-       unsigned long flags;
-
-       spin_lock_irqsave(&spppq_lock, flags);
-
-       for (sp=spppq; sp; sp=sp->pp_next) 
-       {
-               struct net_device *dev = sp->pp_if;
-
-               /* Keepalive mode disabled or channel down? */
-               if (! (sp->pp_flags & PP_KEEPALIVE) ||
-                   ! (dev->flags & IFF_UP))
-                       continue;
-
-               spin_lock(&sp->lock);
-
-               /* No keepalive in PPP mode if LCP not opened yet. */
-               if (! (sp->pp_flags & PP_CISCO) &&
-                   sp->lcp.state != LCP_STATE_OPENED) {
-                       spin_unlock(&sp->lock);
-                       continue;
-               }
-
-               if (sp->pp_alivecnt == MAXALIVECNT) {
-                       /* No keepalive packets got.  Stop the interface. */
-                       printk (KERN_WARNING "%s: protocol down\n", dev->name);
-                       if_down (dev);
-                       if (! (sp->pp_flags & PP_CISCO)) {
-                               /* Shut down the PPP link. */
-                               sp->lcp.magic = jiffies;
-                               sp->lcp.state = LCP_STATE_CLOSED;
-                               sp->ipcp.state = IPCP_STATE_CLOSED;
-                               sppp_clear_timeout (sp);
-                               /* Initiate negotiation. */
-                               sppp_lcp_open (sp);
-                       }
-               }
-               if (sp->pp_alivecnt <= MAXALIVECNT)
-                       ++sp->pp_alivecnt;
-               if (sp->pp_flags & PP_CISCO)
-                       sppp_cisco_send (sp, CISCO_KEEPALIVE_REQ, ++sp->pp_seq,
-                               sp->pp_rseq);
-               else if (sp->lcp.state == LCP_STATE_OPENED) {
-                       __be32 nmagic = htonl (sp->lcp.magic);
-                       sp->lcp.echoid = ++sp->pp_seq;
-                       sppp_cp_send (sp, PPP_LCP, LCP_ECHO_REQ,
-                               sp->lcp.echoid, 4, &nmagic);
-               }
-
-               spin_unlock(&sp->lock);
-       }
-       spin_unlock_irqrestore(&spppq_lock, flags);
-       sppp_flush_xmit();
-       sppp_keepalive_timer.expires=jiffies+10*HZ;
-       add_timer(&sppp_keepalive_timer);
-}
-
-/*
- * Handle incoming PPP Link Control Protocol packets.
- */
-static void sppp_lcp_input (struct sppp *sp, struct sk_buff *skb)
-{
-       struct lcp_header *h;
-       struct net_device *dev = sp->pp_if;
-       int len = skb->len;
-       u8 *p, opt[6];
-       u32 rmagic = 0;
-
-       if (!pskb_may_pull(skb, sizeof(struct lcp_header))) {
-               if (sp->pp_flags & PP_DEBUG)
-                       printk (KERN_WARNING "%s: invalid lcp packet length: %d bytes\n",
-                               dev->name, len);
-               return;
-       }
-       h = (struct lcp_header *)skb->data;
-       skb_pull(skb,sizeof(struct lcp_header *));
-       
-       if (sp->pp_flags & PP_DEBUG) 
-       {
-               char state = '?';
-               switch (sp->lcp.state) {
-               case LCP_STATE_CLOSED:   state = 'C'; break;
-               case LCP_STATE_ACK_RCVD: state = 'R'; break;
-               case LCP_STATE_ACK_SENT: state = 'S'; break;
-               case LCP_STATE_OPENED:   state = 'O'; break;
-               }
-               printk (KERN_WARNING "%s: lcp input(%c): %d bytes <%s id=%xh len=%xh",
-                       dev->name, state, len,
-                       sppp_lcp_type_name (h->type), h->ident, ntohs (h->len));
-               if (len > 4)
-                       sppp_print_bytes ((u8*) (h+1), len-4);
-               printk (">\n");
-       }
-       if (len > ntohs (h->len))
-               len = ntohs (h->len);
-       switch (h->type) {
-       default:
-               /* Unknown packet type -- send Code-Reject packet. */
-               sppp_cp_send (sp, PPP_LCP, LCP_CODE_REJ, ++sp->pp_seq,
-                       skb->len, h);
-               break;
-       case LCP_CONF_REQ:
-               if (len < 4) {
-                       if (sp->pp_flags & PP_DEBUG)
-                               printk (KERN_DEBUG"%s: invalid lcp configure request packet length: %d bytes\n",
-                                       dev->name, len);
-                       break;
-               }
-               if (len>4 && !sppp_lcp_conf_parse_options (sp, h, len, &rmagic))
-                       goto badreq;
-               if (rmagic == sp->lcp.magic) {
-                       /* Local and remote magics equal -- loopback? */
-                       if (sp->pp_loopcnt >= MAXALIVECNT*5) {
-                               printk (KERN_WARNING "%s: loopback\n",
-                                       dev->name);
-                               sp->pp_loopcnt = 0;
-                               if (dev->flags & IFF_UP) {
-                                       if_down (dev);
-                               }
-                       } else if (sp->pp_flags & PP_DEBUG)
-                               printk (KERN_DEBUG "%s: conf req: magic glitch\n",
-                                       dev->name);
-                       ++sp->pp_loopcnt;
-
-                       /* MUST send Conf-Nack packet. */
-                       rmagic = ~sp->lcp.magic;
-                       opt[0] = LCP_OPT_MAGIC;
-                       opt[1] = sizeof (opt);
-                       opt[2] = rmagic >> 24;
-                       opt[3] = rmagic >> 16;
-                       opt[4] = rmagic >> 8;
-                       opt[5] = rmagic;
-                       sppp_cp_send (sp, PPP_LCP, LCP_CONF_NAK,
-                               h->ident, sizeof (opt), &opt);
-badreq:
-                       switch (sp->lcp.state) {
-                       case LCP_STATE_OPENED:
-                               /* Initiate renegotiation. */
-                               sppp_lcp_open (sp);
-                               /* fall through... */
-                       case LCP_STATE_ACK_SENT:
-                               /* Go to closed state. */
-                               sp->lcp.state = LCP_STATE_CLOSED;
-                               sp->ipcp.state = IPCP_STATE_CLOSED;
-                       }
-                       break;
-               }
-               /* Send Configure-Ack packet. */
-               sp->pp_loopcnt = 0;
-               if (sp->lcp.state != LCP_STATE_OPENED) {
-                       sppp_cp_send (sp, PPP_LCP, LCP_CONF_ACK,
-                                       h->ident, len-4, h+1);
-               }
-               /* Change the state. */
-               switch (sp->lcp.state) {
-               case LCP_STATE_CLOSED:
-                       sp->lcp.state = LCP_STATE_ACK_SENT;
-                       break;
-               case LCP_STATE_ACK_RCVD:
-                       sp->lcp.state = LCP_STATE_OPENED;
-                       sppp_ipcp_open (sp);
-                       break;
-               case LCP_STATE_OPENED:
-                       /* Remote magic changed -- close session. */
-                       sp->lcp.state = LCP_STATE_CLOSED;
-                       sp->ipcp.state = IPCP_STATE_CLOSED;
-                       /* Initiate renegotiation. */
-                       sppp_lcp_open (sp);
-                       /* Send ACK after our REQ in attempt to break loop */
-                       sppp_cp_send (sp, PPP_LCP, LCP_CONF_ACK,
-                                       h->ident, len-4, h+1);
-                       sp->lcp.state = LCP_STATE_ACK_SENT;
-                       break;
-               }
-               break;
-       case LCP_CONF_ACK:
-               if (h->ident != sp->lcp.confid)
-                       break;
-               sppp_clear_timeout (sp);
-               if ((sp->pp_link_state != SPPP_LINK_UP) &&
-                   (dev->flags & IFF_UP)) {
-                       /* Coming out of loopback mode. */
-                       sp->pp_link_state=SPPP_LINK_UP;
-                       printk (KERN_INFO "%s: protocol up\n", dev->name);
-               }
-               switch (sp->lcp.state) {
-               case LCP_STATE_CLOSED:
-                       sp->lcp.state = LCP_STATE_ACK_RCVD;
-                       sppp_set_timeout (sp, 5);
-                       break;
-               case LCP_STATE_ACK_SENT:
-                       sp->lcp.state = LCP_STATE_OPENED;
-                       sppp_ipcp_open (sp);
-                       break;
-               }
-               break;
-       case LCP_CONF_NAK:
-               if (h->ident != sp->lcp.confid)
-                       break;
-               p = (u8*) (h+1);
-               if (len>=10 && p[0] == LCP_OPT_MAGIC && p[1] >= 4) {
-                       rmagic = (u32)p[2] << 24 |
-                               (u32)p[3] << 16 | p[4] << 8 | p[5];
-                       if (rmagic == ~sp->lcp.magic) {
-                               int newmagic;
-                               if (sp->pp_flags & PP_DEBUG)
-                                       printk (KERN_DEBUG "%s: conf nak: magic glitch\n",
-                                               dev->name);
-                               get_random_bytes(&newmagic, sizeof(newmagic));
-                               sp->lcp.magic += newmagic;
-                       } else
-                               sp->lcp.magic = rmagic;
-                       }
-               if (sp->lcp.state != LCP_STATE_ACK_SENT) {
-                       /* Go to closed state. */
-                       sp->lcp.state = LCP_STATE_CLOSED;
-                       sp->ipcp.state = IPCP_STATE_CLOSED;
-               }
-               /* The link will be renegotiated after timeout,
-                * to avoid endless req-nack loop. */
-               sppp_clear_timeout (sp);
-               sppp_set_timeout (sp, 2);
-               break;
-       case LCP_CONF_REJ:
-               if (h->ident != sp->lcp.confid)
-                       break;
-               sppp_clear_timeout (sp);
-               /* Initiate renegotiation. */
-               sppp_lcp_open (sp);
-               if (sp->lcp.state != LCP_STATE_ACK_SENT) {
-                       /* Go to closed state. */
-                       sp->lcp.state = LCP_STATE_CLOSED;
-                       sp->ipcp.state = IPCP_STATE_CLOSED;
-               }
-               break;
-       case LCP_TERM_REQ:
-               sppp_clear_timeout (sp);
-               /* Send Terminate-Ack packet. */
-               sppp_cp_send (sp, PPP_LCP, LCP_TERM_ACK, h->ident, 0, NULL);
-               /* Go to closed state. */
-               sp->lcp.state = LCP_STATE_CLOSED;
-               sp->ipcp.state = IPCP_STATE_CLOSED;
-               /* Initiate renegotiation. */
-               sppp_lcp_open (sp);
-               break;
-       case LCP_TERM_ACK:
-       case LCP_CODE_REJ:
-       case LCP_PROTO_REJ:
-               /* Ignore for now. */
-               break;
-       case LCP_DISC_REQ:
-               /* Discard the packet. */
-               break;
-       case LCP_ECHO_REQ:
-               if (sp->lcp.state != LCP_STATE_OPENED)
-                       break;
-               if (len < 8) {
-                       if (sp->pp_flags & PP_DEBUG)
-                               printk (KERN_WARNING "%s: invalid lcp echo request packet length: %d bytes\n",
-                                       dev->name, len);
-                       break;
-               }
-               if (ntohl (*(__be32*)(h+1)) == sp->lcp.magic) {
-                       /* Line loopback mode detected. */
-                       printk (KERN_WARNING "%s: loopback\n", dev->name);
-                       if_down (dev);
-
-                       /* Shut down the PPP link. */
-                       sp->lcp.state = LCP_STATE_CLOSED;
-                       sp->ipcp.state = IPCP_STATE_CLOSED;
-                       sppp_clear_timeout (sp);
-                       /* Initiate negotiation. */
-                       sppp_lcp_open (sp);
-                       break;
-               }
-               *(__be32 *)(h+1) = htonl (sp->lcp.magic);
-               sppp_cp_send (sp, PPP_LCP, LCP_ECHO_REPLY, h->ident, len-4, h+1);
-               break;
-       case LCP_ECHO_REPLY:
-               if (h->ident != sp->lcp.echoid)
-                       break;
-               if (len < 8) {
-                       if (sp->pp_flags & PP_DEBUG)
-                               printk (KERN_WARNING "%s: invalid lcp echo reply packet length: %d bytes\n",
-                                       dev->name, len);
-                       break;
-               }
-               if (ntohl(*(__be32 *)(h+1)) != sp->lcp.magic)
-               sp->pp_alivecnt = 0;
-               break;
-       }
-}
-
-/*
- * Handle incoming Cisco keepalive protocol packets.
- */
-
-static void sppp_cisco_input (struct sppp *sp, struct sk_buff *skb)
-{
-       struct cisco_packet *h;
-       struct net_device *dev = sp->pp_if;
-
-       if (!pskb_may_pull(skb, sizeof(struct cisco_packet))
-           || (skb->len != CISCO_PACKET_LEN
-               && skb->len != CISCO_BIG_PACKET_LEN)) {
-               if (sp->pp_flags & PP_DEBUG)
-                       printk (KERN_WARNING "%s: invalid cisco packet length: %d bytes\n",
-                               dev->name,  skb->len);
-               return;
-       }
-       h = (struct cisco_packet *)skb->data;
-       skb_pull(skb, sizeof(struct cisco_packet*));
-       if (sp->pp_flags & PP_DEBUG)
-               printk (KERN_WARNING "%s: cisco input: %d bytes <%xh %xh %xh %xh %xh-%xh>\n",
-                       dev->name,  skb->len,
-                       ntohl (h->type), h->par1, h->par2, h->rel,
-                       h->time0, h->time1);
-       switch (ntohl (h->type)) {
-       default:
-               if (sp->pp_flags & PP_DEBUG)
-                       printk (KERN_WARNING "%s: unknown cisco packet type: 0x%x\n",
-                               dev->name,  ntohl (h->type));
-               break;
-       case CISCO_ADDR_REPLY:
-               /* Reply on address request, ignore */
-               break;
-       case CISCO_KEEPALIVE_REQ:
-               sp->pp_alivecnt = 0;
-               sp->pp_rseq = ntohl (h->par1);
-               if (sp->pp_seq == sp->pp_rseq) {
-                       /* Local and remote sequence numbers are equal.
-                        * Probably, the line is in loopback mode. */
-                       int newseq;
-                       if (sp->pp_loopcnt >= MAXALIVECNT) {
-                               printk (KERN_WARNING "%s: loopback\n",
-                                       dev->name);
-                               sp->pp_loopcnt = 0;
-                               if (dev->flags & IFF_UP) {
-                                       if_down (dev);
-                               }
-                       }
-                       ++sp->pp_loopcnt;
-
-                       /* Generate new local sequence number */
-                       get_random_bytes(&newseq, sizeof(newseq));
-                       sp->pp_seq ^= newseq;
-                       break;
-               }
-               sp->pp_loopcnt = 0;
-               if (sp->pp_link_state==SPPP_LINK_DOWN &&
-                   (dev->flags & IFF_UP)) {
-                       sp->pp_link_state=SPPP_LINK_UP;
-                       printk (KERN_INFO "%s: protocol up\n", dev->name);
-               }
-               break;
-       case CISCO_ADDR_REQ:
-               /* Stolen from net/ipv4/devinet.c -- SIOCGIFADDR ioctl */
-               {
-               __be32 addr = 0, mask = htonl(~0U); /* FIXME: is the mask correct? */
-#ifdef CONFIG_INET
-               struct in_device *in_dev;
-               struct in_ifaddr *ifa;
-
-               rcu_read_lock();
-               if ((in_dev = __in_dev_get_rcu(dev)) != NULL)
-               {
-                       for (ifa=in_dev->ifa_list; ifa != NULL;
-                               ifa=ifa->ifa_next) {
-                               if (strcmp(dev->name, ifa->ifa_label) == 0) 
-                               {
-                                       addr = ifa->ifa_local;
-                                       mask = ifa->ifa_mask;
-                                       break;
-                               }
-                       }
-               }
-               rcu_read_unlock();
-#endif         
-               sppp_cisco_send (sp, CISCO_ADDR_REPLY, ntohl(addr), ntohl(mask));
-               break;
-               }
-       }
-}
-
-
-/*
- * Send PPP LCP packet.
- */
-
-static void sppp_cp_send (struct sppp *sp, u16 proto, u8 type,
-       u8 ident, u16 len, void *data)
-{
-       struct ppp_header *h;
-       struct lcp_header *lh;
-       struct sk_buff *skb;
-       struct net_device *dev = sp->pp_if;
-
-       skb=alloc_skb(dev->hard_header_len+PPP_HEADER_LEN+LCP_HEADER_LEN+len,
-               GFP_ATOMIC);
-       if (skb==NULL)
-               return;
-
-       skb_reserve(skb,dev->hard_header_len);
-       
-       h = (struct ppp_header *)skb_put(skb, sizeof(struct ppp_header));
-       h->address = PPP_ALLSTATIONS;        /* broadcast address */
-       h->control = PPP_UI;                 /* Unnumbered Info */
-       h->protocol = htons (proto);         /* Link Control Protocol */
-
-       lh = (struct lcp_header *)skb_put(skb, sizeof(struct lcp_header));
-       lh->type = type;
-       lh->ident = ident;
-       lh->len = htons (LCP_HEADER_LEN + len);
-
-       if (len)
-               memcpy(skb_put(skb,len),data, len);
-
-       if (sp->pp_flags & PP_DEBUG) {
-               printk (KERN_WARNING "%s: %s output <%s id=%xh len=%xh",
-                       dev->name, 
-                       proto==PPP_LCP ? "lcp" : "ipcp",
-                       proto==PPP_LCP ? sppp_lcp_type_name (lh->type) :
-                       sppp_ipcp_type_name (lh->type), lh->ident,
-                       ntohs (lh->len));
-               if (len)
-                       sppp_print_bytes ((u8*) (lh+1), len);
-               printk (">\n");
-       }
-       /* Control is high priority so it doesn't get queued behind data */
-       skb->priority=TC_PRIO_CONTROL;
-       skb->dev = dev;
-       skb_queue_tail(&tx_queue, skb);
-}
-
-/*
- * Send Cisco keepalive packet.
- */
-
-static void sppp_cisco_send (struct sppp *sp, int type, u32 par1, u32 par2)
-{
-       struct ppp_header *h;
-       struct cisco_packet *ch;
-       struct sk_buff *skb;
-       struct net_device *dev = sp->pp_if;
-       u32 t = jiffies * 1000/HZ;
-
-       skb=alloc_skb(dev->hard_header_len+PPP_HEADER_LEN+CISCO_PACKET_LEN,
-               GFP_ATOMIC);
-
-       if(skb==NULL)
-               return;
-               
-       skb_reserve(skb, dev->hard_header_len);
-       h = (struct ppp_header *)skb_put (skb, sizeof(struct ppp_header));
-       h->address = CISCO_MULTICAST;
-       h->control = 0;
-       h->protocol = htons (CISCO_KEEPALIVE);
-
-       ch = (struct cisco_packet*)skb_put(skb, CISCO_PACKET_LEN);
-       ch->type = htonl (type);
-       ch->par1 = htonl (par1);
-       ch->par2 = htonl (par2);
-       ch->rel = htons(0xffff);
-       ch->time0 = htons ((u16) (t >> 16));
-       ch->time1 = htons ((u16) t);
-
-       if (sp->pp_flags & PP_DEBUG)
-               printk (KERN_WARNING "%s: cisco output: <%xh %xh %xh %xh %xh-%xh>\n",
-                       dev->name,  ntohl (ch->type), ch->par1,
-                       ch->par2, ch->rel, ch->time0, ch->time1);
-       skb->priority=TC_PRIO_CONTROL;
-       skb->dev = dev;
-       skb_queue_tail(&tx_queue, skb);
-}
-
-/**
- *     sppp_close - close down a synchronous PPP or Cisco HDLC link
- *     @dev: The network device to drop the link of
- *
- *     This drops the logical interface to the channel. It is not
- *     done politely as we assume we will also be dropping DTR. Any
- *     timeouts are killed.
- */
-
-int sppp_close (struct net_device *dev)
-{
-       struct sppp *sp = (struct sppp *)sppp_of(dev);
-       unsigned long flags;
-
-       spin_lock_irqsave(&sp->lock, flags);
-       sp->pp_link_state = SPPP_LINK_DOWN;
-       sp->lcp.state = LCP_STATE_CLOSED;
-       sp->ipcp.state = IPCP_STATE_CLOSED;
-       sppp_clear_timeout (sp);
-       spin_unlock_irqrestore(&sp->lock, flags);
-
-       return 0;
-}
-
-EXPORT_SYMBOL(sppp_close);
-
-/**
- *     sppp_open - open a synchronous PPP or Cisco HDLC link
- *     @dev:   Network device to activate
- *     
- *     Close down any existing synchronous session and commence
- *     from scratch. In the PPP case this means negotiating LCP/IPCP
- *     and friends, while for Cisco HDLC we simply need to start sending
- *     keepalives
- */
-
-int sppp_open (struct net_device *dev)
-{
-       struct sppp *sp = (struct sppp *)sppp_of(dev);
-       unsigned long flags;
-
-       sppp_close(dev);
-
-       spin_lock_irqsave(&sp->lock, flags);
-       if (!(sp->pp_flags & PP_CISCO)) {
-               sppp_lcp_open (sp);
-       }
-       sp->pp_link_state = SPPP_LINK_DOWN;
-       spin_unlock_irqrestore(&sp->lock, flags);
-       sppp_flush_xmit();
-
-       return 0;
-}
-
-EXPORT_SYMBOL(sppp_open);
-
-/**
- *     sppp_reopen - notify of physical link loss
- *     @dev: Device that lost the link
- *
- *     This function informs the synchronous protocol code that
- *     the underlying link died (for example a carrier drop on X.21)
- *
- *     We increment the magic numbers to ensure that if the other end
- *     failed to notice we will correctly start a new session. It happens
- *     do to the nature of telco circuits is that you can lose carrier on
- *     one endonly.
- *
- *     Having done this we go back to negotiating. This function may
- *     be called from an interrupt context.
- */
-int sppp_reopen (struct net_device *dev)
-{
-       struct sppp *sp = (struct sppp *)sppp_of(dev);
-       unsigned long flags;
-
-       sppp_close(dev);
-
-       spin_lock_irqsave(&sp->lock, flags);
-       if (!(sp->pp_flags & PP_CISCO))
-       {
-               sp->lcp.magic = jiffies;
-               ++sp->pp_seq;
-               sp->lcp.state = LCP_STATE_CLOSED;
-               sp->ipcp.state = IPCP_STATE_CLOSED;
-               /* Give it a moment for the line to settle then go */
-               sppp_set_timeout (sp, 1);
-       } 
-       sp->pp_link_state=SPPP_LINK_DOWN;
-       spin_unlock_irqrestore(&sp->lock, flags);
-
-       return 0;
-}
-
-EXPORT_SYMBOL(sppp_reopen);
-
-/**
- *     sppp_change_mtu - Change the link MTU
- *     @dev:   Device to change MTU on
- *     @new_mtu: New MTU
- *
- *     Change the MTU on the link. This can only be called with
- *     the link down. It returns an error if the link is up or
- *     the mtu is out of range.
- */
-static int sppp_change_mtu(struct net_device *dev, int new_mtu)
-{
-       if(new_mtu<128||new_mtu>PPP_MTU||(dev->flags&IFF_UP))
-               return -EINVAL;
-       dev->mtu=new_mtu;
-       return 0;
-}
-
-/**
- *     sppp_do_ioctl - Ioctl handler for ppp/hdlc
- *     @dev: Device subject to ioctl
- *     @ifr: Interface request block from the user
- *     @cmd: Command that is being issued
- *     
- *     This function handles the ioctls that may be issued by the user
- *     to control the settings of a PPP/HDLC link. It does both busy
- *     and security checks. This function is intended to be wrapped by
- *     callers who wish to add additional ioctl calls of their own.
- */
-int sppp_do_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
-{
-       struct sppp *sp = (struct sppp *)sppp_of(dev);
-
-       if(dev->flags&IFF_UP)
-               return -EBUSY;
-               
-       if(!capable(CAP_NET_ADMIN))
-               return -EPERM;
-       
-       switch(cmd)
-       {
-               case SPPPIOCCISCO:
-                       sp->pp_flags|=PP_CISCO;
-                       dev->type = ARPHRD_HDLC;
-                       break;
-               case SPPPIOCPPP:
-                       sp->pp_flags&=~PP_CISCO;
-                       dev->type = ARPHRD_PPP;
-                       break;
-               case SPPPIOCDEBUG:
-                       sp->pp_flags&=~PP_DEBUG;
-                       if(ifr->ifr_flags)
-                               sp->pp_flags|=PP_DEBUG;
-                       break;
-               case SPPPIOCGFLAGS:
-                       if(copy_to_user(ifr->ifr_data, &sp->pp_flags, sizeof(sp->pp_flags)))
-                               return -EFAULT;
-                       break;
-               case SPPPIOCSFLAGS:
-                       if(copy_from_user(&sp->pp_flags, ifr->ifr_data, sizeof(sp->pp_flags)))
-                               return -EFAULT;
-                       break;
-               default:
-                       return -EINVAL;
-       }
-       return 0;
-}
-
-EXPORT_SYMBOL(sppp_do_ioctl);
-
-/**
- *     sppp_attach - attach synchronous PPP/HDLC to a device
- *     @pd:    PPP device to initialise
- *
- *     This initialises the PPP/HDLC support on an interface. At the
- *     time of calling the dev element must point to the network device
- *     that this interface is attached to. The interface should not yet
- *     be registered. 
- */
-void sppp_attach(struct ppp_device *pd)
-{
-       struct net_device *dev = pd->dev;
-       struct sppp *sp = &pd->sppp;
-       unsigned long flags;
-
-       /* Make sure embedding is safe for sppp_of */
-       BUG_ON(sppp_of(dev) != sp);
-
-       spin_lock_irqsave(&spppq_lock, flags);
-       /* Initialize keepalive handler. */
-       if (! spppq)
-       {
-               init_timer(&sppp_keepalive_timer);
-               sppp_keepalive_timer.expires=jiffies+10*HZ;
-               sppp_keepalive_timer.function=sppp_keepalive;
-               add_timer(&sppp_keepalive_timer);
-       }
-       /* Insert new entry into the keepalive list. */
-       sp->pp_next = spppq;
-       spppq = sp;
-       spin_unlock_irqrestore(&spppq_lock, flags);
-
-       sp->pp_loopcnt = 0;
-       sp->pp_alivecnt = 0;
-       sp->pp_seq = 0;
-       sp->pp_rseq = 0;
-       sp->pp_flags = PP_KEEPALIVE|PP_CISCO|debug;/*PP_DEBUG;*/
-       sp->lcp.magic = 0;
-       sp->lcp.state = LCP_STATE_CLOSED;
-       sp->ipcp.state = IPCP_STATE_CLOSED;
-       sp->pp_if = dev;
-       spin_lock_init(&sp->lock);
-       
-       /* 
-        *      Device specific setup. All but interrupt handler and
-        *      hard_start_xmit.
-        */
-        
-       dev->header_ops = &sppp_header_ops;
-
-       dev->tx_queue_len = 10;
-       dev->type = ARPHRD_HDLC;
-       dev->addr_len = 0;
-       dev->hard_header_len = sizeof(struct ppp_header);
-       dev->mtu = PPP_MTU;
-       /*
-        *      These 4 are callers but MUST also call sppp_ functions
-        */
-       dev->do_ioctl = sppp_do_ioctl;
-#if 0
-       dev->get_stats = NULL;          /* Let the driver override these */
-       dev->open = sppp_open;
-       dev->stop = sppp_close;
-#endif 
-       dev->change_mtu = sppp_change_mtu;
-       dev->flags = IFF_MULTICAST|IFF_POINTOPOINT|IFF_NOARP;
-}
-
-EXPORT_SYMBOL(sppp_attach);
-
-/**
- *     sppp_detach - release PPP resources from a device
- *     @dev:   Network device to release
- *
- *     Stop and free up any PPP/HDLC resources used by this
- *     interface. This must be called before the device is
- *     freed.
- */
-void sppp_detach (struct net_device *dev)
-{
-       struct sppp **q, *p, *sp = (struct sppp *)sppp_of(dev);
-       unsigned long flags;
-
-       spin_lock_irqsave(&spppq_lock, flags);
-       /* Remove the entry from the keepalive list. */
-       for (q = &spppq; (p = *q); q = &p->pp_next)
-               if (p == sp) {
-                       *q = p->pp_next;
-                       break;
-               }
-
-       /* Stop keepalive handler. */
-       if (! spppq)
-               del_timer(&sppp_keepalive_timer);
-       sppp_clear_timeout (sp);
-       spin_unlock_irqrestore(&spppq_lock, flags);
-}
-
-EXPORT_SYMBOL(sppp_detach);
-
-/*
- * Analyze the LCP Configure-Request options list
- * for the presence of unknown options.
- * If the request contains unknown options, build and
- * send Configure-reject packet, containing only unknown options.
- */
-static int
-sppp_lcp_conf_parse_options (struct sppp *sp, struct lcp_header *h,
-       int len, u32 *magic)
-{
-       u8 *buf, *r, *p;
-       int rlen;
-
-       len -= 4;
-       buf = r = kmalloc (len, GFP_ATOMIC);
-       if (! buf)
-               return (0);
-
-       p = (void*) (h+1);
-       for (rlen=0; len>1 && p[1]; len-=p[1], p+=p[1]) {
-               switch (*p) {
-               case LCP_OPT_MAGIC:
-                       /* Magic number -- extract. */
-                       if (len >= 6 && p[1] == 6) {
-                               *magic = (u32)p[2] << 24 |
-                                       (u32)p[3] << 16 | p[4] << 8 | p[5];
-                               continue;
-                       }
-                       break;
-               case LCP_OPT_ASYNC_MAP:
-                       /* Async control character map -- check to be zero. */
-                       if (len >= 6 && p[1] == 6 && ! p[2] && ! p[3] &&
-                           ! p[4] && ! p[5])
-                               continue;
-                       break;
-               case LCP_OPT_MRU:
-                       /* Maximum receive unit -- always OK. */
-                       continue;
-               default:
-                       /* Others not supported. */
-                       break;
-               }
-               /* Add the option to rejected list. */
-               memcpy(r, p, p[1]);
-               r += p[1];
-               rlen += p[1];
-       }
-       if (rlen)
-               sppp_cp_send (sp, PPP_LCP, LCP_CONF_REJ, h->ident, rlen, buf);
-       kfree(buf);
-       return (rlen == 0);
-}
-
-static void sppp_ipcp_input (struct sppp *sp, struct sk_buff *skb)
-{
-       struct lcp_header *h;
-       struct net_device *dev = sp->pp_if;
-       int len = skb->len;
-
-       if (!pskb_may_pull(skb, sizeof(struct lcp_header))) {
-               if (sp->pp_flags & PP_DEBUG)
-                       printk (KERN_WARNING "%s: invalid ipcp packet length: %d bytes\n",
-                               dev->name,  len);
-               return;
-       }
-       h = (struct lcp_header *)skb->data;
-       skb_pull(skb,sizeof(struct lcp_header));
-       if (sp->pp_flags & PP_DEBUG) {
-               printk (KERN_WARNING "%s: ipcp input: %d bytes <%s id=%xh len=%xh",
-                       dev->name,  len,
-                       sppp_ipcp_type_name (h->type), h->ident, ntohs (h->len));
-               if (len > 4)
-                       sppp_print_bytes ((u8*) (h+1), len-4);
-               printk (">\n");
-       }
-       if (len > ntohs (h->len))
-               len = ntohs (h->len);
-       switch (h->type) {
-       default:
-               /* Unknown packet type -- send Code-Reject packet. */
-               sppp_cp_send (sp, PPP_IPCP, IPCP_CODE_REJ, ++sp->pp_seq, len, h);
-               break;
-       case IPCP_CONF_REQ:
-               if (len < 4) {
-                       if (sp->pp_flags & PP_DEBUG)
-                               printk (KERN_WARNING "%s: invalid ipcp configure request packet length: %d bytes\n",
-                                       dev->name, len);
-                       return;
-               }
-               if (len > 4) {
-                       sppp_cp_send (sp, PPP_IPCP, LCP_CONF_REJ, h->ident,
-                               len-4, h+1);
-
-                       switch (sp->ipcp.state) {
-                       case IPCP_STATE_OPENED:
-                               /* Initiate renegotiation. */
-                               sppp_ipcp_open (sp);
-                               /* fall through... */
-                       case IPCP_STATE_ACK_SENT:
-                               /* Go to closed state. */
-                               sp->ipcp.state = IPCP_STATE_CLOSED;
-                       }
-               } else {
-                       /* Send Configure-Ack packet. */
-                       sppp_cp_send (sp, PPP_IPCP, IPCP_CONF_ACK, h->ident,
-                               0, NULL);
-                       /* Change the state. */
-                       if (sp->ipcp.state == IPCP_STATE_ACK_RCVD)
-                               sp->ipcp.state = IPCP_STATE_OPENED;
-                       else
-                               sp->ipcp.state = IPCP_STATE_ACK_SENT;
-               }
-               break;
-       case IPCP_CONF_ACK:
-               if (h->ident != sp->ipcp.confid)
-                       break;
-               sppp_clear_timeout (sp);
-               switch (sp->ipcp.state) {
-               case IPCP_STATE_CLOSED:
-                       sp->ipcp.state = IPCP_STATE_ACK_RCVD;
-                       sppp_set_timeout (sp, 5);
-                       break;
-               case IPCP_STATE_ACK_SENT:
-                       sp->ipcp.state = IPCP_STATE_OPENED;
-                       break;
-               }
-               break;
-       case IPCP_CONF_NAK:
-       case IPCP_CONF_REJ:
-               if (h->ident != sp->ipcp.confid)
-                       break;
-               sppp_clear_timeout (sp);
-                       /* Initiate renegotiation. */
-               sppp_ipcp_open (sp);
-               if (sp->ipcp.state != IPCP_STATE_ACK_SENT)
-                       /* Go to closed state. */
-                       sp->ipcp.state = IPCP_STATE_CLOSED;
-               break;
-       case IPCP_TERM_REQ:
-               /* Send Terminate-Ack packet. */
-               sppp_cp_send (sp, PPP_IPCP, IPCP_TERM_ACK, h->ident, 0, NULL);
-               /* Go to closed state. */
-               sp->ipcp.state = IPCP_STATE_CLOSED;
-               /* Initiate renegotiation. */
-               sppp_ipcp_open (sp);
-               break;
-       case IPCP_TERM_ACK:
-               /* Ignore for now. */
-       case IPCP_CODE_REJ:
-               /* Ignore for now. */
-               break;
-       }
-}
-
-static void sppp_lcp_open (struct sppp *sp)
-{
-       char opt[6];
-
-       if (! sp->lcp.magic)
-               sp->lcp.magic = jiffies;
-       opt[0] = LCP_OPT_MAGIC;
-       opt[1] = sizeof (opt);
-       opt[2] = sp->lcp.magic >> 24;
-       opt[3] = sp->lcp.magic >> 16;
-       opt[4] = sp->lcp.magic >> 8;
-       opt[5] = sp->lcp.magic;
-       sp->lcp.confid = ++sp->pp_seq;
-       sppp_cp_send (sp, PPP_LCP, LCP_CONF_REQ, sp->lcp.confid,
-               sizeof (opt), &opt);
-       sppp_set_timeout (sp, 2);
-}
-
-static void sppp_ipcp_open (struct sppp *sp)
-{
-       sp->ipcp.confid = ++sp->pp_seq;
-       sppp_cp_send (sp, PPP_IPCP, IPCP_CONF_REQ, sp->ipcp.confid, 0, NULL);
-       sppp_set_timeout (sp, 2);
-}
-
-/*
- * Process PPP control protocol timeouts.
- */
-static void sppp_cp_timeout (unsigned long arg)
-{
-       struct sppp *sp = (struct sppp*) arg;
-       unsigned long flags;
-
-       spin_lock_irqsave(&sp->lock, flags);
-
-       sp->pp_flags &= ~PP_TIMO;
-       if (! (sp->pp_if->flags & IFF_UP) || (sp->pp_flags & PP_CISCO)) {
-               spin_unlock_irqrestore(&sp->lock, flags);
-               return;
-       }
-       switch (sp->lcp.state) {
-       case LCP_STATE_CLOSED:
-               /* No ACK for Configure-Request, retry. */
-               sppp_lcp_open (sp);
-               break;
-       case LCP_STATE_ACK_RCVD:
-               /* ACK got, but no Configure-Request for peer, retry. */
-               sppp_lcp_open (sp);
-               sp->lcp.state = LCP_STATE_CLOSED;
-               break;
-       case LCP_STATE_ACK_SENT:
-               /* ACK sent but no ACK for Configure-Request, retry. */
-               sppp_lcp_open (sp);
-               break;
-       case LCP_STATE_OPENED:
-               /* LCP is already OK, try IPCP. */
-               switch (sp->ipcp.state) {
-               case IPCP_STATE_CLOSED:
-                       /* No ACK for Configure-Request, retry. */
-                       sppp_ipcp_open (sp);
-                       break;
-               case IPCP_STATE_ACK_RCVD:
-                       /* ACK got, but no Configure-Request for peer, retry. */
-                       sppp_ipcp_open (sp);
-                       sp->ipcp.state = IPCP_STATE_CLOSED;
-                       break;
-               case IPCP_STATE_ACK_SENT:
-                       /* ACK sent but no ACK for Configure-Request, retry. */
-                       sppp_ipcp_open (sp);
-                       break;
-               case IPCP_STATE_OPENED:
-                       /* IPCP is OK. */
-                       break;
-               }
-               break;
-       }
-       spin_unlock_irqrestore(&sp->lock, flags);
-       sppp_flush_xmit();
-}
-
-static char *sppp_lcp_type_name (u8 type)
-{
-       static char buf [8];
-       switch (type) {
-       case LCP_CONF_REQ:   return ("conf-req");
-       case LCP_CONF_ACK:   return ("conf-ack");
-       case LCP_CONF_NAK:   return ("conf-nack");
-       case LCP_CONF_REJ:   return ("conf-rej");
-       case LCP_TERM_REQ:   return ("term-req");
-       case LCP_TERM_ACK:   return ("term-ack");
-       case LCP_CODE_REJ:   return ("code-rej");
-       case LCP_PROTO_REJ:  return ("proto-rej");
-       case LCP_ECHO_REQ:   return ("echo-req");
-       case LCP_ECHO_REPLY: return ("echo-reply");
-       case LCP_DISC_REQ:   return ("discard-req");
-       }
-       sprintf (buf, "%xh", type);
-       return (buf);
-}
-
-static char *sppp_ipcp_type_name (u8 type)
-{
-       static char buf [8];
-       switch (type) {
-       case IPCP_CONF_REQ:   return ("conf-req");
-       case IPCP_CONF_ACK:   return ("conf-ack");
-       case IPCP_CONF_NAK:   return ("conf-nack");
-       case IPCP_CONF_REJ:   return ("conf-rej");
-       case IPCP_TERM_REQ:   return ("term-req");
-       case IPCP_TERM_ACK:   return ("term-ack");
-       case IPCP_CODE_REJ:   return ("code-rej");
-       }
-       sprintf (buf, "%xh", type);
-       return (buf);
-}
-
-static void sppp_print_bytes (u_char *p, u16 len)
-{
-       printk (" %x", *p++);
-       while (--len > 0)
-               printk ("-%x", *p++);
-}
-
-/**
- *     sppp_rcv -      receive and process a WAN PPP frame
- *     @skb:   The buffer to process
- *     @dev:   The device it arrived on
- *     @p: Unused
- *     @orig_dev: Unused
- *
- *     Protocol glue. This drives the deferred processing mode the poorer
- *     cards use. This can be called directly by cards that do not have
- *     timing constraints but is normally called from the network layer
- *     after interrupt servicing to process frames queued via netif_rx.
- */
-
-static int sppp_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *p, struct net_device *orig_dev)
-{
-       if (dev_net(dev) != &init_net) {
-               kfree_skb(skb);
-               return 0;
-       }
-
-       if ((skb = skb_share_check(skb, GFP_ATOMIC)) == NULL)
-               return NET_RX_DROP;
-       sppp_input(dev,skb);
-       return 0;
-}
-
-static struct packet_type sppp_packet_type = {
-       .type   = __constant_htons(ETH_P_WAN_PPP),
-       .func   = sppp_rcv,
-};
-
-static char banner[] __initdata = 
-       KERN_INFO "Cronyx Ltd, Synchronous PPP and CISCO HDLC (c) 1994\n"
-       KERN_INFO "Linux port (c) 1998 Building Number Three Ltd & "
-                 "Jan \"Yenya\" Kasprzak.\n";
-
-static int __init sync_ppp_init(void)
-{
-       if(debug)
-               debug=PP_DEBUG;
-       printk(banner);
-       skb_queue_head_init(&tx_queue);
-       dev_add_pack(&sppp_packet_type);
-       return 0;
-}
-
-
-static void __exit sync_ppp_cleanup(void)
-{
-       dev_remove_pack(&sppp_packet_type);
-}
-
-module_init(sync_ppp_init);
-module_exit(sync_ppp_cleanup);
-module_param(debug, int, 0);
-MODULE_LICENSE("GPL");
-
diff --git a/include/net/syncppp.h b/include/net/syncppp.h
deleted file mode 100644 (file)
index 9e306f7..0000000
+++ /dev/null
@@ -1,102 +0,0 @@
-/*
- * Defines for synchronous PPP/Cisco link level subroutines.
- *
- * Copyright (C) 1994 Cronyx Ltd.
- * Author: Serge Vakulenko, <vak@zebub.msk.su>
- *
- * This software is distributed with NO WARRANTIES, not even the implied
- * warranties for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
- *
- * Authors grant any other persons or organizations permission to use
- * or modify this software as long as this message is kept with the software,
- * all derivative works or modified versions.
- *
- * Version 1.7, Wed Jun  7 22:12:02 MSD 1995
- *
- *
- *
- */
-
-#ifndef _SYNCPPP_H_
-#define _SYNCPPP_H_ 1
-
-#ifdef __KERNEL__
-struct slcp {
-       u16     state;          /* state machine */
-       u32     magic;          /* local magic number */
-       u_char  echoid;         /* id of last keepalive echo request */
-       u_char  confid;         /* id of last configuration request */
-};
-
-struct sipcp {
-       u16     state;          /* state machine */
-       u_char  confid;         /* id of last configuration request */
-};
-
-struct sppp 
-{
-       struct sppp *   pp_next;        /* next interface in keepalive list */
-       u32             pp_flags;       /* use Cisco protocol instead of PPP */
-       u16             pp_alivecnt;    /* keepalive packets counter */
-       u16             pp_loopcnt;     /* loopback detection counter */
-       u32             pp_seq;         /* local sequence number */
-       u32             pp_rseq;        /* remote sequence number */
-       struct slcp     lcp;            /* LCP params */
-       struct sipcp    ipcp;           /* IPCP params */
-       struct timer_list       pp_timer;
-       struct net_device       *pp_if;
-       char            pp_link_state;  /* Link status */
-       spinlock_t      lock;
-};
-
-struct ppp_device
-{      
-       struct net_device *dev; /* Network device pointer */
-       struct sppp sppp;       /* Synchronous PPP */
-};
-
-static inline struct sppp *sppp_of(struct net_device *dev) 
-{
-       struct ppp_device **ppp = dev->ml_priv;
-       BUG_ON((*ppp)->dev != dev);
-       return &(*ppp)->sppp;
-}
-
-#define PP_KEEPALIVE    0x01    /* use keepalive protocol */
-#define PP_CISCO        0x02    /* use Cisco protocol instead of PPP */
-#define PP_TIMO         0x04    /* cp_timeout routine active */
-#define PP_DEBUG       0x08
-
-#define PPP_MTU          1500    /* max. transmit unit */
-
-#define LCP_STATE_CLOSED        0       /* LCP state: closed (conf-req sent) */
-#define LCP_STATE_ACK_RCVD      1       /* LCP state: conf-ack received */
-#define LCP_STATE_ACK_SENT      2       /* LCP state: conf-ack sent */
-#define LCP_STATE_OPENED        3       /* LCP state: opened */
-
-#define IPCP_STATE_CLOSED       0       /* IPCP state: closed (conf-req sent) */
-#define IPCP_STATE_ACK_RCVD     1       /* IPCP state: conf-ack received */
-#define IPCP_STATE_ACK_SENT     2       /* IPCP state: conf-ack sent */
-#define IPCP_STATE_OPENED       3       /* IPCP state: opened */
-
-#define SPPP_LINK_DOWN         0       /* link down - no keepalive */
-#define SPPP_LINK_UP           1       /* link is up - keepalive ok */
-
-void sppp_attach (struct ppp_device *pd);
-void sppp_detach (struct net_device *dev);
-int sppp_do_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd);
-struct sk_buff *sppp_dequeue (struct net_device *dev);
-int sppp_isempty (struct net_device *dev);
-void sppp_flush (struct net_device *dev);
-int sppp_open (struct net_device *dev);
-int sppp_reopen (struct net_device *dev);
-int sppp_close (struct net_device *dev);
-#endif
-
-#define SPPPIOCCISCO   (SIOCDEVPRIVATE)
-#define SPPPIOCPPP     (SIOCDEVPRIVATE+1)
-#define SPPPIOCDEBUG   (SIOCDEVPRIVATE+2)
-#define SPPPIOCSFLAGS  (SIOCDEVPRIVATE+3)
-#define SPPPIOCGFLAGS  (SIOCDEVPRIVATE+4)
-
-#endif /* _SYNCPPP_H_ */