d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
030e4ebba8
freebsd-skq/usr.sbin/ppp/lcp.c

731 lines
19 KiB
C
Raw Normal View History

This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
/*
* PPP Link Control Protocol (LCP) Module
*
* Written by Toshiharu OHNO (tony-o@iij.ad.jp)
*
* Copyright (C) 1993, Internet Initiative Japan, Inc. All rights reserverd.
*
* Redistribution and use in source and binary forms are permitted
* provided that the above copyright notice and this paragraph are
* duplicated in all such forms and that any documentation,
* advertising materials, and other materials related to such
* distribution and use acknowledge that the software was developed
* by the Internet Initiative Japan, Inc. The name of the
* IIJ may not be used to endorse or promote products derived
* from this software without specific prior written permission.
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
* WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE.
*
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
* $Id: lcp.c,v 1.55.2.41 1998/04/07 00:53:55 brian Exp $
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
*
* TODO:
* o Limit data field length by MRU
*/
#include <signal.h>
#include <stdio.h>
#include <stdlib.h>
#include <termios.h>
#include <unistd.h>
#include "command.h"
#include "mbuf.h"
#include "log.h"
#include "defs.h"
#include "timer.h"
#include "fsm.h"
#include "lcp.h"
o Move global/static IPCP data into IpcpInfo. o Use INADDR_ANY and INADDR_BROADCAST instead of 0 and 0xffffffff for IP addresses & masks. o Move struct compreq into ipcp.c (it's none of anyone elses business). o Rename MAX_STATES to MAX_VJ_STATES.
1998-01-29 23:11:44 +00:00
#include "throughput.h"
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#include "lcpproto.h"
Make struct bundle into a sort of `struct descriptor'. It does the fdsets/reads/writes for each of it's datalinks.
1998-03-20 19:48:28 +00:00
#include "descriptor.h"
De-staticise LQR information Increment OutPackets for any packet - not just LQRs MFC: o Fix a few comment typos. o Fix ``set timeout'' usage message and documentation. o Change ifOutPackets, ifOutOctets and ifOutLQRs to `u_int32_t's so that they wrap correctly. o Put the LQR in network byte order using the correct struct size (sizeof u_int32_t, not sizeof u_long). o Wrap LQR ECHO counters correctly. o Don't increment OutLQR count if the last LQR hasn't been replied to. o Initialise last received LQR in StartLqm. o Don't start the LQR timer if we're `disabled' and `accepted'. o Generate LQR responses when both sides are using a timer and we're not going to send our next LQR before the peers max timeout.
1998-03-13 00:44:26 +00:00
#include "lqr.h"
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#include "hdlc.h"
#include "ccp.h"
#include "async.h"
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
#include "link.h"
Create a new MP branch for `multilink protocol'. Do lots of initial shuffling and grouping. Submitted by: Eivind Eklund <perhaps@yes.no>
1998-01-29 00:49:32 +00:00
#include "physical.h"
Move the terminal/diagnostic socket stuff out of main.c and into the new `prompt.c'. struct prompt is (of course) a `sort' of descriptor.
1998-02-10 03:23:50 +00:00
#include "prompt.h"
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
/* for received LQRs */
struct lqrreq {
u_char type;
u_char length;
u_short proto; /* Quality protocol */
u_long period; /* Reporting interval */
};
static void LcpLayerUp(struct fsm *);
static void LcpLayerDown(struct fsm *);
static void LcpLayerStart(struct fsm *);
static void LcpLayerFinish(struct fsm *);
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
static void LcpInitRestartCounter(struct fsm *);
static void LcpSendConfigReq(struct fsm *);
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
static void LcpSentTerminateReq(struct fsm *);
static void LcpSendTerminateAck(struct fsm *, u_char);
Remove more globals
1998-03-13 21:08:05 +00:00
static void LcpDecodeConfig(struct fsm *, u_char *, int, int,
struct fsm_decode *);
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
static struct fsm_callbacks lcp_Callbacks = {
Create struct fsm_parent. FSMs are created with one of these, and the FSM passes subsequent events to them. The datalink now hides its CCP from the bundle layer.
1998-02-27 01:22:39 +00:00
LcpLayerUp,
LcpLayerDown,
LcpLayerStart,
LcpLayerFinish,
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
LcpInitRestartCounter,
LcpSendConfigReq,
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
LcpSentTerminateReq,
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
LcpSendTerminateAck,
Shuffle around our FSMs a bit. This'll make it easier to remove the CcpInfo, LcpInfo and IpcpInfo globals.
1998-02-21 01:45:26 +00:00
LcpDecodeConfig,
NullRecvResetReq,
NullRecvResetAck
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
};
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
static const char *cftypes[] = {
/* Check out the latest ``Assigned numbers'' rfc (rfc1700.txt) */
"???",
"MRU", /* 1: Maximum-Receive-Unit */
"ACCMAP", /* 2: Async-Control-Character-Map */
"AUTHPROTO", /* 3: Authentication-Protocol */
"QUALPROTO", /* 4: Quality-Protocol */
"MAGICNUM", /* 5: Magic-Number */
"RESERVED", /* 6: RESERVED */
"PROTOCOMP", /* 7: Protocol-Field-Compression */
"ACFCOMP", /* 8: Address-and-Control-Field-Compression */
"FCSALT", /* 9: FCS-Alternatives */
"SDP", /* 10: Self-Describing-Pad */
"NUMMODE", /* 11: Numbered-Mode */
"MULTIPROC", /* 12: Multi-Link-Procedure */
"CALLBACK", /* 13: Callback */
"CONTIME", /* 14: Connect-Time */
"COMPFRAME", /* 15: Compound-Frames */
"NDE", /* 16: Nominal-Data-Encapsulation */
"MULTIMRRU", /* 17: Multilink-MRRU */
"MULTISSNH", /* 18: Multilink-Short-Sequence-Number-Header */
"MULTIED", /* 19: Multilink-Endpoint-Descriminator */
"PROPRIETRY", /* 20: Proprietary */
"DCEID", /* 21: DCE-Identifier */
"MULTIPP", /* 22: Multi-Link-Plus-Procedure */
"LDBACP", /* 23: Link Discriminator for BACP */
};
#define NCFTYPES (sizeof cftypes/sizeof cftypes[0])
int
o Don't immediately reply to the first LQR thinking it's a duplicate ('cos it compares with our initial values). o Move the LCP into struct datalink.
1998-03-13 21:07:14 +00:00
lcp_ReportStatus(struct cmdargs const *arg)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
o Move default MRU, MTU, ACCMAP and OPENMODE config values into struct lcp and display them in `show lcp'. o Remove `show mru' and `show mtu' and make the data part of `show lcp'. Also merge `set m[tr]u' and `set openmode' implementations into the SetVariable function. o `set timeout' only accepts the idle timer value as an argument. o Move our lqr period into struct lcp, and create a `set lqrperiod' command. Display it in `show lcp'. o Remove VarRetryTimeout, and implement it at the LCP, PAP, CHAP, CCP and IPCP levels, creating individual `set XXXretry' commands for each. They must be separate because they have different context requirements in multilink mode. o Display default config values in `show ccp'. o Tart the man page up a bit (wrt PPP/TCP, compression and LQR) and explain the new commands.
1998-04-03 19:24:07 +00:00
struct link *l = ChooseLink(arg);
struct lcp *lcp = &l->lcp;
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
Deglobalise `struct prompt': o Our diagnostic socket has its password set in the `set socket' line only (not in ppp.secret). o Passwords are per server socket (*VarAuthKey are gone) o Authority is per prompt (VarLocalAuth is gone). o Local logging is per prompt. o Add a `show who' command to see who's connected. No identd routine - just a `where the connection came from' display. o SIGUSR1 is disabled for now - we have no way of choosing a password for the socket created :-( Prompts are attached as a list of `struct descriptor's in struct bundle, and serviced under the bundles descriptor service routines. Ultimately, everything should be done like this. Cosmetic: o alphabeticalise SRCS in Makefile. o Add a few comments in command.h TODO: Start checking that we don't overflow the descriptor sets in select() now that we can have any number of descriptors.
1998-04-03 19:26:02 +00:00
prompt_Printf(arg->prompt, "%s: %s [%s]\n", l->name, lcp->fsm.name,
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
State2Nam(lcp->fsm.state));
Deglobalise `struct prompt': o Our diagnostic socket has its password set in the `set socket' line only (not in ppp.secret). o Passwords are per server socket (*VarAuthKey are gone) o Authority is per prompt (VarLocalAuth is gone). o Local logging is per prompt. o Add a `show who' command to see who's connected. No identd routine - just a `where the connection came from' display. o SIGUSR1 is disabled for now - we have no way of choosing a password for the socket created :-( Prompts are attached as a list of `struct descriptor's in struct bundle, and serviced under the bundles descriptor service routines. Ultimately, everything should be done like this. Cosmetic: o alphabeticalise SRCS in Makefile. o Add a few comments in command.h TODO: Start checking that we don't overflow the descriptor sets in select() now that we can have any number of descriptors.
1998-04-03 19:26:02 +00:00
prompt_Printf(arg->prompt,
`int var : 1' -> `unsigned var : 1'. This made things look silly and made us mis-negotiate the number of VJ slots.
1998-04-05 19:02:49 +00:00
" his side: MRU %d, ACCMAP %08lx, PROTOCOMP %s, ACFCOMP %s,\n"
Move the terminal/diagnostic socket stuff out of main.c and into the new `prompt.c'. struct prompt is (of course) a `sort' of descriptor.
1998-02-10 03:23:50 +00:00
" MAGIC %08lx, REJECT %04x\n",
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
lcp->his_mru, (u_long)lcp->his_accmap,
`int var : 1' -> `unsigned var : 1'. This made things look silly and made us mis-negotiate the number of VJ slots.
1998-04-05 19:02:49 +00:00
lcp->his_protocomp ? "on" : "off",
lcp->his_acfcomp ? "on" : "off",
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
(u_long)lcp->his_magic, lcp->his_reject);
Deglobalise `struct prompt': o Our diagnostic socket has its password set in the `set socket' line only (not in ppp.secret). o Passwords are per server socket (*VarAuthKey are gone) o Authority is per prompt (VarLocalAuth is gone). o Local logging is per prompt. o Add a `show who' command to see who's connected. No identd routine - just a `where the connection came from' display. o SIGUSR1 is disabled for now - we have no way of choosing a password for the socket created :-( Prompts are attached as a list of `struct descriptor's in struct bundle, and serviced under the bundles descriptor service routines. Ultimately, everything should be done like this. Cosmetic: o alphabeticalise SRCS in Makefile. o Add a few comments in command.h TODO: Start checking that we don't overflow the descriptor sets in select() now that we can have any number of descriptors.
1998-04-03 19:26:02 +00:00
prompt_Printf(arg->prompt,
`int var : 1' -> `unsigned var : 1'. This made things look silly and made us mis-negotiate the number of VJ slots.
1998-04-05 19:02:49 +00:00
" my side: MRU %d, ACCMAP %08lx, PROTOCOMP %s, ACFCOMP %s,\n"
Move the terminal/diagnostic socket stuff out of main.c and into the new `prompt.c'. struct prompt is (of course) a `sort' of descriptor.
1998-02-10 03:23:50 +00:00
" MAGIC %08lx, REJECT %04x\n",
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
lcp->want_mru, (u_long)lcp->want_accmap,
`int var : 1' -> `unsigned var : 1'. This made things look silly and made us mis-negotiate the number of VJ slots.
1998-04-05 19:02:49 +00:00
lcp->want_protocomp ? "on" : "off",
lcp->want_acfcomp ? "on" : "off",
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
(u_long)lcp->want_magic, lcp->my_reject);
o Move default MRU, MTU, ACCMAP and OPENMODE config values into struct lcp and display them in `show lcp'. o Remove `show mru' and `show mtu' and make the data part of `show lcp'. Also merge `set m[tr]u' and `set openmode' implementations into the SetVariable function. o `set timeout' only accepts the idle timer value as an argument. o Move our lqr period into struct lcp, and create a `set lqrperiod' command. Display it in `show lcp'. o Remove VarRetryTimeout, and implement it at the LCP, PAP, CHAP, CCP and IPCP levels, creating individual `set XXXretry' commands for each. They must be separate because they have different context requirements in multilink mode. o Display default config values in `show ccp'. o Tart the man page up a bit (wrt PPP/TCP, compression and LQR) and explain the new commands.
1998-04-03 19:24:07 +00:00
Deglobalise `struct prompt': o Our diagnostic socket has its password set in the `set socket' line only (not in ppp.secret). o Passwords are per server socket (*VarAuthKey are gone) o Authority is per prompt (VarLocalAuth is gone). o Local logging is per prompt. o Add a `show who' command to see who's connected. No identd routine - just a `where the connection came from' display. o SIGUSR1 is disabled for now - we have no way of choosing a password for the socket created :-( Prompts are attached as a list of `struct descriptor's in struct bundle, and serviced under the bundles descriptor service routines. Ultimately, everything should be done like this. Cosmetic: o alphabeticalise SRCS in Makefile. o Add a few comments in command.h TODO: Start checking that we don't overflow the descriptor sets in select() now that we can have any number of descriptors.
1998-04-03 19:26:02 +00:00
prompt_Printf(arg->prompt, "\n Defaults: MRU = %d, ", lcp->cfg.mru);
o Move default MRU, MTU, ACCMAP and OPENMODE config values into struct lcp and display them in `show lcp'. o Remove `show mru' and `show mtu' and make the data part of `show lcp'. Also merge `set m[tr]u' and `set openmode' implementations into the SetVariable function. o `set timeout' only accepts the idle timer value as an argument. o Move our lqr period into struct lcp, and create a `set lqrperiod' command. Display it in `show lcp'. o Remove VarRetryTimeout, and implement it at the LCP, PAP, CHAP, CCP and IPCP levels, creating individual `set XXXretry' commands for each. They must be separate because they have different context requirements in multilink mode. o Display default config values in `show ccp'. o Tart the man page up a bit (wrt PPP/TCP, compression and LQR) and explain the new commands.
1998-04-03 19:24:07 +00:00
if (l->lcp.cfg.mtu)
Deglobalise `struct prompt': o Our diagnostic socket has its password set in the `set socket' line only (not in ppp.secret). o Passwords are per server socket (*VarAuthKey are gone) o Authority is per prompt (VarLocalAuth is gone). o Local logging is per prompt. o Add a `show who' command to see who's connected. No identd routine - just a `where the connection came from' display. o SIGUSR1 is disabled for now - we have no way of choosing a password for the socket created :-( Prompts are attached as a list of `struct descriptor's in struct bundle, and serviced under the bundles descriptor service routines. Ultimately, everything should be done like this. Cosmetic: o alphabeticalise SRCS in Makefile. o Add a few comments in command.h TODO: Start checking that we don't overflow the descriptor sets in select() now that we can have any number of descriptors.
1998-04-03 19:26:02 +00:00
prompt_Printf(arg->prompt, "MTU = %d, ", lcp->cfg.mtu);
prompt_Printf(arg->prompt, "ACCMAP = %08lx\n", (u_long)lcp->cfg.accmap);
prompt_Printf(arg->prompt, " LQR period = %us, ",
lcp->cfg.lqrperiod);
prompt_Printf(arg->prompt, "Open Mode = %s",
o Move default MRU, MTU, ACCMAP and OPENMODE config values into struct lcp and display them in `show lcp'. o Remove `show mru' and `show mtu' and make the data part of `show lcp'. Also merge `set m[tr]u' and `set openmode' implementations into the SetVariable function. o `set timeout' only accepts the idle timer value as an argument. o Move our lqr period into struct lcp, and create a `set lqrperiod' command. Display it in `show lcp'. o Remove VarRetryTimeout, and implement it at the LCP, PAP, CHAP, CCP and IPCP levels, creating individual `set XXXretry' commands for each. They must be separate because they have different context requirements in multilink mode. o Display default config values in `show ccp'. o Tart the man page up a bit (wrt PPP/TCP, compression and LQR) and explain the new commands.
1998-04-03 19:24:07 +00:00
lcp->cfg.openmode == OPEN_PASSIVE ? "passive" : "active");
if (lcp->cfg.openmode > 0)
Deglobalise `struct prompt': o Our diagnostic socket has its password set in the `set socket' line only (not in ppp.secret). o Passwords are per server socket (*VarAuthKey are gone) o Authority is per prompt (VarLocalAuth is gone). o Local logging is per prompt. o Add a `show who' command to see who's connected. No identd routine - just a `where the connection came from' display. o SIGUSR1 is disabled for now - we have no way of choosing a password for the socket created :-( Prompts are attached as a list of `struct descriptor's in struct bundle, and serviced under the bundles descriptor service routines. Ultimately, everything should be done like this. Cosmetic: o alphabeticalise SRCS in Makefile. o Add a few comments in command.h TODO: Start checking that we don't overflow the descriptor sets in select() now that we can have any number of descriptors.
1998-04-03 19:26:02 +00:00
prompt_Printf(arg->prompt, " (delay %ds)", lcp->cfg.openmode);
prompt_Printf(arg->prompt, "\n FSM retry = %us\n",
lcp->cfg.fsmretry);
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
prompt_Printf(arg->prompt, "\n Negotiation:\n");
prompt_Printf(arg->prompt, " ACFCOMP = %s\n",
command_ShowNegval(lcp->cfg.acfcomp));
prompt_Printf(arg->prompt, " CHAP = %s\n",
command_ShowNegval(lcp->cfg.chap));
prompt_Printf(arg->prompt, " LQR = %s\n",
command_ShowNegval(lcp->cfg.lqr));
prompt_Printf(arg->prompt, " PAP = %s\n",
command_ShowNegval(lcp->cfg.pap));
prompt_Printf(arg->prompt, " PROTOCOMP = %s\n",
command_ShowNegval(lcp->cfg.protocomp));
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
return 0;
}
static u_int32_t
GenerateMagic(void)
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* Generate random number which will be used as magic number */
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
randinit();
return (random());
}
void
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
lcp_Init(struct lcp *lcp, struct bundle *bundle, struct link *l,
Create struct fsm_parent. FSMs are created with one of these, and the FSM passes subsequent events to them. The datalink now hides its CCP from the bundle layer.
1998-02-27 01:22:39 +00:00
const struct fsm_parent *parent)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* Initialise ourselves */
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
int mincode = parent ? 1 : LCP_MINMPCODE;
static const char *timer_names[] =
{"LCP restart", "LCP openmode", "LCP stopped"};
fsm_Init(&lcp->fsm, "LCP", PROTO_LCP, mincode, LCP_MAXCODE, 10, LogLCP,
bundle, l, parent, &lcp_Callbacks, timer_names);
o Move default MRU, MTU, ACCMAP and OPENMODE config values into struct lcp and display them in `show lcp'. o Remove `show mru' and `show mtu' and make the data part of `show lcp'. Also merge `set m[tr]u' and `set openmode' implementations into the SetVariable function. o `set timeout' only accepts the idle timer value as an argument. o Move our lqr period into struct lcp, and create a `set lqrperiod' command. Display it in `show lcp'. o Remove VarRetryTimeout, and implement it at the LCP, PAP, CHAP, CCP and IPCP levels, creating individual `set XXXretry' commands for each. They must be separate because they have different context requirements in multilink mode. o Display default config values in `show ccp'. o Tart the man page up a bit (wrt PPP/TCP, compression and LQR) and explain the new commands.
1998-04-03 19:24:07 +00:00
lcp->cfg.mru = DEF_MRU;
lcp->cfg.mtu = DEF_MTU;
lcp->cfg.accmap = 0;
lcp->cfg.openmode = 1;
lcp->cfg.lqrperiod = DEF_LQRPERIOD;
lcp->cfg.fsmretry = DEF_FSMRETRY;
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
lcp->cfg.acfcomp = NEG_ENABLED|NEG_ACCEPTED;
lcp->cfg.chap = NEG_ACCEPTED;
lcp->cfg.lqr = NEG_ACCEPTED;
lcp->cfg.pap = NEG_ACCEPTED;
lcp->cfg.protocomp = NEG_ENABLED|NEG_ACCEPTED;
o Move default MRU, MTU, ACCMAP and OPENMODE config values into struct lcp and display them in `show lcp'. o Remove `show mru' and `show mtu' and make the data part of `show lcp'. Also merge `set m[tr]u' and `set openmode' implementations into the SetVariable function. o `set timeout' only accepts the idle timer value as an argument. o Move our lqr period into struct lcp, and create a `set lqrperiod' command. Display it in `show lcp'. o Remove VarRetryTimeout, and implement it at the LCP, PAP, CHAP, CCP and IPCP levels, creating individual `set XXXretry' commands for each. They must be separate because they have different context requirements in multilink mode. o Display default config values in `show ccp'. o Tart the man page up a bit (wrt PPP/TCP, compression and LQR) and explain the new commands.
1998-04-03 19:24:07 +00:00
lcp_Setup(lcp, lcp->cfg.openmode);
Shuffle around our FSMs a bit. This'll make it easier to remove the CcpInfo, LcpInfo and IpcpInfo globals.
1998-02-21 01:45:26 +00:00
}
void
lcp_Setup(struct lcp *lcp, int openmode)
{
lcp->fsm.open_mode = openmode;
Reset fsm.maxconfig in the *_Setup() routines.
1998-02-27 21:46:26 +00:00
lcp->fsm.maxconfig = 10;
Shuffle around our FSMs a bit. This'll make it easier to remove the CcpInfo, LcpInfo and IpcpInfo globals.
1998-02-21 01:45:26 +00:00
lcp->his_mru = DEF_MRU;
lcp->his_magic = 0;
lcp->his_lqrperiod = 0;
lcp->his_acfcomp = 0;
lcp->his_auth = 0;
o Move default MRU, MTU, ACCMAP and OPENMODE config values into struct lcp and display them in `show lcp'. o Remove `show mru' and `show mtu' and make the data part of `show lcp'. Also merge `set m[tr]u' and `set openmode' implementations into the SetVariable function. o `set timeout' only accepts the idle timer value as an argument. o Move our lqr period into struct lcp, and create a `set lqrperiod' command. Display it in `show lcp'. o Remove VarRetryTimeout, and implement it at the LCP, PAP, CHAP, CCP and IPCP levels, creating individual `set XXXretry' commands for each. They must be separate because they have different context requirements in multilink mode. o Display default config values in `show ccp'. o Tart the man page up a bit (wrt PPP/TCP, compression and LQR) and explain the new commands.
1998-04-03 19:24:07 +00:00
lcp->want_mru = lcp->cfg.mru;
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
lcp->want_acfcomp = IsEnabled(lcp->cfg.acfcomp) ? 1 : 0;
Shuffle around our FSMs a bit. This'll make it easier to remove the CcpInfo, LcpInfo and IpcpInfo globals.
1998-02-21 01:45:26 +00:00
Initialise multilink LCPs according to rfc1990.
1998-04-03 19:26:12 +00:00
if (lcp->fsm.parent) {
lcp->his_accmap = 0xffffffff;
lcp->want_accmap = lcp->cfg.accmap;
lcp->his_protocomp = 0;
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
lcp->want_protocomp = IsEnabled(lcp->cfg.protocomp) ? 1 : 0;
Initialise multilink LCPs according to rfc1990.
1998-04-03 19:26:12 +00:00
lcp->want_magic = GenerateMagic();
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
lcp->want_auth = IsEnabled(lcp->cfg.chap) ? PROTO_CHAP :
IsEnabled(lcp->cfg.pap) ? PROTO_PAP : 0;
lcp->want_lqrperiod = IsEnabled(lcp->cfg.lqr) ?
lcp->cfg.lqrperiod * 100 : 0;
Initialise multilink LCPs according to rfc1990.
1998-04-03 19:26:12 +00:00
} else {
lcp->his_accmap = lcp->want_accmap = 0;
lcp->his_protocomp = lcp->want_protocomp = 1;
lcp->want_magic = 0;
lcp->want_auth = 0;
lcp->want_lqrperiod = 0;
}
Shuffle around our FSMs a bit. This'll make it easier to remove the CcpInfo, LcpInfo and IpcpInfo globals.
1998-02-21 01:45:26 +00:00
lcp->his_reject = lcp->my_reject = 0;
lcp->auth_iwait = lcp->auth_ineed = 0;
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
lcp->LcpFailedMagic = 0;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
static void
LcpInitRestartCounter(struct fsm * fp)
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* Set fsm timer load */
o Move default MRU, MTU, ACCMAP and OPENMODE config values into struct lcp and display them in `show lcp'. o Remove `show mru' and `show mtu' and make the data part of `show lcp'. Also merge `set m[tr]u' and `set openmode' implementations into the SetVariable function. o `set timeout' only accepts the idle timer value as an argument. o Move our lqr period into struct lcp, and create a `set lqrperiod' command. Display it in `show lcp'. o Remove VarRetryTimeout, and implement it at the LCP, PAP, CHAP, CCP and IPCP levels, creating individual `set XXXretry' commands for each. They must be separate because they have different context requirements in multilink mode. o Display default config values in `show ccp'. o Tart the man page up a bit (wrt PPP/TCP, compression and LQR) and explain the new commands.
1998-04-03 19:24:07 +00:00
struct lcp *lcp = fsm2lcp(fp);
fp->FsmTimer.load = lcp->cfg.fsmretry * SECTICKS;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
fp->restart = 5;
}
static void
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
LcpSendConfigReq(struct fsm *fp)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* Send config REQ please */
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
struct physical *p = link2physical(fp->link);
Rename ccpstate to ccp, ipcpstate to ipcp and lcpstate to lcp.
1998-02-08 11:05:01 +00:00
struct lcp *lcp = fsm2lcp(fp);
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
u_char buff[100];
struct lcp_opt *o;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
if (!p) {
LogPrintf(LogERROR, "LcpSendConfigReq: Not a physical link !\n");
return;
}
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
o = (struct lcp_opt *)buff;
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
if (!Physical_IsSync(p)) {
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
if (lcp->want_acfcomp && !REJECTED(lcp, TY_ACFCOMP))
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
INC_LCP_OPT(TY_ACFCOMP, 2, o);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
if (lcp->want_protocomp && !REJECTED(lcp, TY_PROTOCOMP))
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
INC_LCP_OPT(TY_PROTOCOMP, 2, o);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
if (!REJECTED(lcp, TY_ACCMAP)) {
*(u_int32_t *)o->data = htonl(lcp->want_accmap);
INC_LCP_OPT(TY_ACCMAP, 6, o);
}
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
if (!REJECTED(lcp, TY_MRU)) {
*(u_short *)o->data = htons(lcp->want_mru);
INC_LCP_OPT(TY_MRU, 4, o);
}
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
if (lcp->want_magic && !REJECTED(lcp, TY_MAGICNUM)) {
*(u_int32_t *)o->data = htonl(lcp->want_magic);
INC_LCP_OPT(TY_MAGICNUM, 6, o);
}
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
if (lcp->want_lqrperiod && !REJECTED(lcp, TY_QUALPROTO)) {
*(u_short *)o->data = htons(PROTO_LQR);
*(u_long *)(o->data + 2) = htonl(lcp->want_lqrperiod);
INC_LCP_OPT(TY_QUALPROTO, 8, o);
}
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
switch (lcp->want_auth) {
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
case PROTO_PAP:
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
*(u_short *)o->data = htons(PROTO_PAP);
INC_LCP_OPT(TY_AUTHPROTO, 4, o);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
case PROTO_CHAP:
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
*(u_short *)o->data = htons(PROTO_CHAP);
o Move VarMSChap into struct chap, and remove `set encrypt' option. We never ask a client for MSChap when we've got chap `enabled', and we dynamically answer using MSChap if the peer demands it. o Remove all of the bundle2*() series of functions except bundle2datalink() as they're too expensive. The only calls to bundle2datalink() are made from command.c when determining context. o Write to the correct modem in term mode, and check the return value, dropping back to command mode if the write fails. Cosmetic: Make the PPP COMMAND LIST section of the man page prettier, better and more consistent. Alphabeticalise all command lists and document missing commands.
1998-04-03 19:26:29 +00:00
o->data[2] = 0x05;
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
INC_LCP_OPT(TY_AUTHPROTO, 5, o);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
FsmOutput(fp, CODE_CONFIGREQ, fp->reqid, buff, (u_char *)o - buff);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
void
o Don't immediately reply to the first LQR thinking it's a duplicate ('cos it compares with our initial values). o Move the LCP into struct datalink.
1998-03-13 21:07:14 +00:00
lcp_SendProtoRej(struct lcp *lcp, u_char *option, int count)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* Don't understand `option' */
o Don't immediately reply to the first LQR thinking it's a duplicate ('cos it compares with our initial values). o Move the LCP into struct datalink.
1998-03-13 21:07:14 +00:00
FsmOutput(&lcp->fsm, CODE_PROTOREJ, lcp->fsm.reqid, option, count);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
static void
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
LcpSentTerminateReq(struct fsm * fp)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* Term REQ just sent by FSM */
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
static void
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
LcpSendTerminateAck(struct fsm *fp, u_char id)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* Send Term ACK please */
o Check the LCP Identifier field for correctness and drop dodgy packets by default. The old behaviour is still available with ``disable idcheck''. o Make all FSM diagnostics consistent and tidy up the way we build our LCP/CCP/IPCP requests. o Don't assume sizeof(u_long) == 4.
1998-03-20 19:47:10 +00:00
FsmOutput(fp, CODE_TERMACK, id, NULL, 0);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
static void
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
LcpLayerStart(struct fsm *fp)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* We're about to start up ! */
o Don't immediately reply to the first LQR thinking it's a duplicate ('cos it compares with our initial values). o Move the LCP into struct datalink.
1998-03-13 21:07:14 +00:00
struct lcp *lcp = fsm2lcp(fp);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
LogPrintf(LogLCP, "LcpLayerStart\n");
o Don't immediately reply to the first LQR thinking it's a duplicate ('cos it compares with our initial values). o Move the LCP into struct datalink.
1998-03-13 21:07:14 +00:00
lcp->LcpFailedMagic = 0;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
static void
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
LcpLayerFinish(struct fsm *fp)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* We're now down */
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
LogPrintf(LogLCP, "LcpLayerFinish\n");
}
static void
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
LcpLayerUp(struct fsm *fp)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* We're now up */
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
struct physical *p = link2physical(fp->link);
Rename ccpstate to ccp, ipcpstate to ipcp and lcpstate to lcp.
1998-02-08 11:05:01 +00:00
struct lcp *lcp = fsm2lcp(fp);
Create `struct link' - the logical link from which `struct physical' (and the future `struct logical') are derived.
1998-01-30 19:46:07 +00:00
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
LogPrintf(LogLCP, "LcpLayerUp\n");
o Move struct lcp and struct ccp into struct link. o Remove bundle2lcp(), bundle2ccp() and bundle2link(). They're too resource-hungry and we have `owner pointers' to do their job. o Make our FSM understand LCPs that are always ST_OPENED (with a minimum code that != 1). o Send FSM code rejects for invalid codes. o Make our bundle fsm_parent deal with multiple links. o Make timer diagnostics pretty and allow access via ~t in `term' mode (not just when logging debug) and `show timers'. Only show timers every second in debug mode, otherwise we get too many diagnostics to be useful (we probably still do). Also, don't restrict ~m in term mode to depend on debug logging. o Rationalise our bundles' phases. o Create struct mp (multilink protocol). This is both an NCP and a type of struct link. It feeds off other NCPs for output, passing fragmented packets into the queues of available datalinks. It also gets PROTO_MP input, reassembles the fragments into ppp frames, and passes them back to the HDLC layer that the fragments were passed from. ** It's not yet possible to enter multilink mode :-( ** o Add `set weight' (requires context) for deciding on a links weighting in multilink mode. Weighting is simplistic (and probably badly implemented) for now. o Remove the function pointers in struct link. They ended up only applying to physical links. o Configure our tun device with an MTU equal to the MRU from struct mp's LCP and a speed equal to the sum of our link speeds. o `show {lcp,ccp,proto}' and `set deflate' now have optional context and use ChooseLink() to decide on which `struct link' to use. This allows behaviour as before when in non-multilink mode, and allows access to the MP logical link in multilink mode. o Ignore reconnect and redial values when in -direct mode and when cleaning up. Always redial when in -ddial or -dedicated mode (unless cleaning up). o Tell our links to `staydown' when we close them due to a signal. o Remove remaining `#ifdef SIGALRM's (ppp doesn't function without alarms). o Don't bother strdup()ing our physical link name. o Various other cosmetic changes.
1998-04-03 19:21:56 +00:00
async_SetLinkParams(&p->async, lcp);
StartLqm(lcp);
hdlc_StartTimer(&p->hdlc);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
static void
Create `struct bundle' - the top level control structure. This structure will eventually contain a list of NCPs (currently only IPCP is supported) and a list of physical `struct link's. It will also derive from a struct link itself. Make ModemTimeout() static - it's way to dangerous to be called from outside ! Bump version to 1.9. Our first MP release should be 2.0.
1998-02-02 19:32:16 +00:00
LcpLayerDown(struct fsm *fp)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* About to come down */
o Create `struct hdlc' - a part of struct physical. o Move our LCP report timer into struct hdlc - it's really a hdlc timer (fcs errors etc). o Make `show hdlc' require context and make the output more friendly. o Remove all non-const globals from hdlc.c o Output peer-rejected protocols by name - not just ones that we reject.
1998-02-18 19:35:59 +00:00
struct physical *p = link2physical(fp->link);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
LogPrintf(LogLCP, "LcpLayerDown\n");
o Create `struct hdlc' - a part of struct physical. o Move our LCP report timer into struct hdlc - it's really a hdlc timer (fcs errors etc). o Make `show hdlc' require context and make the output more friendly. o Remove all non-const globals from hdlc.c o Output peer-rejected protocols by name - not just ones that we reject.
1998-02-18 19:35:59 +00:00
hdlc_StopTimer(&p->hdlc);
De-staticise LQR information Increment OutPackets for any packet - not just LQRs MFC: o Fix a few comment typos. o Fix ``set timeout'' usage message and documentation. o Change ifOutPackets, ifOutOctets and ifOutLQRs to `u_int32_t's so that they wrap correctly. o Put the LQR in network byte order using the correct struct size (sizeof u_int32_t, not sizeof u_long). o Wrap LQR ECHO counters correctly. o Don't increment OutLQR count if the last LQR hasn't been replied to. o Initialise last received LQR in StartLqm. o Don't start the LQR timer if we're `disabled' and `accepted'. o Generate LQR responses when both sides are using a timer and we're not going to send our next LQR before the peers max timeout.
1998-03-13 00:44:26 +00:00
StopLqrTimer(p);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
static void
Remove more globals
1998-03-13 21:08:05 +00:00
LcpDecodeConfig(struct fsm *fp, u_char *cp, int plen, int mode_type,
struct fsm_decode *dec)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* Deal with incoming PROTO_LCP */
Rename ccpstate to ccp, ipcpstate to ipcp and lcpstate to lcp.
1998-02-08 11:05:01 +00:00
struct lcp *lcp = fsm2lcp(fp);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
int type, length, sz, pos;
u_int32_t *lp, magic, accmap;
u_short mtu, mru, *sp, proto;
struct lqrreq *req;
char request[20], desc[22];
while (plen >= sizeof(struct fsmconfig)) {
type = *cp;
length = cp[1];
if (type < 0 || type >= NCFTYPES)
snprintf(request, sizeof request, " <%d>[%d]", type, length);
else
snprintf(request, sizeof request, " %s[%d]", cftypes[type], length);
switch (type) {
case TY_MRU:
sp = (u_short *) (cp + 2);
mru = htons(*sp);
LogPrintf(LogLCP, "%s %d\n", request, mru);
switch (mode_type) {
case MODE_REQ:
o Move default MRU, MTU, ACCMAP and OPENMODE config values into struct lcp and display them in `show lcp'. o Remove `show mru' and `show mtu' and make the data part of `show lcp'. Also merge `set m[tr]u' and `set openmode' implementations into the SetVariable function. o `set timeout' only accepts the idle timer value as an argument. o Move our lqr period into struct lcp, and create a `set lqrperiod' command. Display it in `show lcp'. o Remove VarRetryTimeout, and implement it at the LCP, PAP, CHAP, CCP and IPCP levels, creating individual `set XXXretry' commands for each. They must be separate because they have different context requirements in multilink mode. o Display default config values in `show ccp'. o Tart the man page up a bit (wrt PPP/TCP, compression and LQR) and explain the new commands.
1998-04-03 19:24:07 +00:00
mtu = lcp->cfg.mtu;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
if (mtu == 0)
mtu = MAX_MTU;
if (mru > mtu) {
*sp = htons(mtu);
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->nakend, cp, 4);
dec->nakend += 4;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
} else if (mru < MIN_MRU) {
*sp = htons(MIN_MRU);
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->nakend, cp, 4);
dec->nakend += 4;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
} else {
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_mru = mru;
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, 4);
dec->ackend += 4;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
break;
case MODE_NAK:
if (mru >= MIN_MRU || mru <= MAX_MRU)
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->want_mru = mru;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
case MODE_REJ:
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
break;
case TY_ACCMAP:
lp = (u_int32_t *) (cp + 2);
accmap = htonl(*lp);
LogPrintf(LogLCP, "%s 0x%08lx\n", request, (u_long)accmap);
switch (mode_type) {
case MODE_REQ:
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_accmap = accmap;
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, 6);
dec->ackend += 6;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
case MODE_NAK:
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->want_accmap = accmap;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
case MODE_REJ:
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
break;
case TY_AUTHPROTO:
sp = (u_short *) (cp + 2);
proto = ntohs(*sp);
switch (proto) {
case PROTO_PAP:
LogPrintf(LogLCP, "%s 0x%04x (PAP)\n", request, proto);
break;
case PROTO_CHAP:
LogPrintf(LogLCP, "%s 0x%04x (CHAP 0x%02x)\n", request, proto, cp[4]);
break;
default:
LogPrintf(LogLCP, "%s 0x%04x\n", request, proto);
break;
}
switch (mode_type) {
case MODE_REQ:
switch (proto) {
case PROTO_PAP:
if (length != 4) {
LogPrintf(LogLCP, " Bad length!\n");
goto reqreject;
}
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
if (IsAccepted(lcp->cfg.pap)) {
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_auth = proto;
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, length);
dec->ackend += length;
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
} else if (IsAccepted(lcp->cfg.chap)) {
Remove more globals
1998-03-13 21:08:05 +00:00
*dec->nakend++ = *cp;
*dec->nakend++ = 5;
*dec->nakend++ = (unsigned char) (PROTO_CHAP >> 8);
*dec->nakend++ = (unsigned char) PROTO_CHAP;
o Move VarMSChap into struct chap, and remove `set encrypt' option. We never ask a client for MSChap when we've got chap `enabled', and we dynamically answer using MSChap if the peer demands it. o Remove all of the bundle2*() series of functions except bundle2datalink() as they're too expensive. The only calls to bundle2datalink() are made from command.c when determining context. o Write to the correct modem in term mode, and check the return value, dropping back to command mode if the write fails. Cosmetic: Make the PPP COMMAND LIST section of the man page prettier, better and more consistent. Alphabeticalise all command lists and document missing commands.
1998-04-03 19:26:29 +00:00
*dec->nakend++ = 0x05;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
} else
goto reqreject;
break;
case PROTO_CHAP:
if (length < 5) {
LogPrintf(LogLCP, " Bad length!\n");
goto reqreject;
}
#ifdef HAVE_DES
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
if (IsAccepted(lcp->cfg.chap) && (cp[4] == 0x05 || cp[4] == 0x80))
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#else
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
if (IsAccepted(lcp->cfg.chap) && cp[4] == 0x05)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#endif
{
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_auth = proto;
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, length);
dec->ackend += length;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#ifdef HAVE_DES
o Move VarMSChap into struct chap, and remove `set encrypt' option. We never ask a client for MSChap when we've got chap `enabled', and we dynamically answer using MSChap if the peer demands it. o Remove all of the bundle2*() series of functions except bundle2datalink() as they're too expensive. The only calls to bundle2datalink() are made from command.c when determining context. o Write to the correct modem in term mode, and check the return value, dropping back to command mode if the write fails. Cosmetic: Make the PPP COMMAND LIST section of the man page prettier, better and more consistent. Alphabeticalise all command lists and document missing commands.
1998-04-03 19:26:29 +00:00
link2physical(fp->link)->dl->chap.using_MSChap = cp[4] == 0x80;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#endif
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
} else if (IsAccepted(lcp->cfg.pap)) {
Remove more globals
1998-03-13 21:08:05 +00:00
*dec->nakend++ = *cp;
*dec->nakend++ = 4;
*dec->nakend++ = (unsigned char) (PROTO_PAP >> 8);
*dec->nakend++ = (unsigned char) PROTO_PAP;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
} else
goto reqreject;
break;
default:
LogPrintf(LogLCP, "%s 0x%04x - not recognised, NAK\n",
request, proto);
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->nakend, cp, length);
dec->nakend += length;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
break;
case MODE_NAK:
switch (proto) {
case PROTO_PAP:
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
if (IsEnabled(lcp->cfg.pap))
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->want_auth = PROTO_PAP;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
else {
LogPrintf(LogLCP, "Peer will only send PAP (not enabled)\n");
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
break;
case PROTO_CHAP:
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
if (IsEnabled(lcp->cfg.chap))
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->want_auth = PROTO_CHAP;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
else {
LogPrintf(LogLCP, "Peer will only send CHAP (not enabled)\n");
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
break;
default:
/* We've been NAK'd with something we don't understand :-( */
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
break;
case MODE_REJ:
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
break;
case TY_QUALPROTO:
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
req = (struct lqrreq *)cp;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
LogPrintf(LogLCP, "%s proto %x, interval %dms\n",
request, ntohs(req->proto), ntohl(req->period) * 10);
switch (mode_type) {
case MODE_REQ:
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
if (ntohs(req->proto) != PROTO_LQR || !IsAccepted(lcp->cfg.lqr))
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
goto reqreject;
else {
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_lqrperiod = ntohl(req->period);
if (lcp->his_lqrperiod < 500)
lcp->his_lqrperiod = 500;
req->period = htonl(lcp->his_lqrperiod);
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, length);
dec->ackend += length;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
break;
case MODE_NAK:
break;
case MODE_REJ:
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
break;
case TY_MAGICNUM:
lp = (u_int32_t *) (cp + 2);
magic = ntohl(*lp);
LogPrintf(LogLCP, "%s 0x%08lx\n", request, (u_long)magic);
switch (mode_type) {
case MODE_REQ:
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
if (lcp->want_magic) {
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
/* Validate magic number */
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
if (magic == lcp->want_magic) {
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
LogPrintf(LogLCP, "Magic is same (%08lx) - %d times\n",
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
(u_long)magic, ++lcp->LcpFailedMagic);
lcp->want_magic = GenerateMagic();
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->nakend, cp, 6);
dec->nakend += 6;
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
ualarm(TICKUNIT * (4 + 4 * lcp->LcpFailedMagic), 0);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
sigpause(0);
} else {
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_magic = magic;
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, length);
dec->ackend += length;
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->LcpFailedMagic = 0;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
} else {
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->my_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
goto reqreject;
}
break;
case MODE_NAK:
LogPrintf(LogLCP, " Magic 0x%08lx is NAKed!\n", (u_long)magic);
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->want_magic = GenerateMagic();
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
case MODE_REJ:
Correct magic number format string Spotted by: Bruce M. Walter <walter@fortean.com>
1998-02-04 01:03:51 +00:00
LogPrintf(LogLCP, " Magic 0x%08x is REJected!\n", magic);
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->want_magic = 0;
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
break;
case TY_PROTOCOMP:
LogPrintf(LogLCP, "%s\n", request);
switch (mode_type) {
case MODE_REQ:
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
if (IsAccepted(lcp->cfg.protocomp)) {
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_protocomp = 1;
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, 2);
dec->ackend += 2;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
} else {
#ifdef OLDMST
/*
* MorningStar before v1.3 needs NAK
*/
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->nakend, cp, 2);
dec->nakend += 2;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#else
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->rejend, cp, 2);
dec->rejend += 2;
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->my_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#endif
}
break;
case MODE_NAK:
case MODE_REJ:
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->want_protocomp = 0;
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
break;
case TY_ACFCOMP:
LogPrintf(LogLCP, "%s\n", request);
switch (mode_type) {
case MODE_REQ:
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
if (IsAccepted(lcp->cfg.acfcomp)) {
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->his_acfcomp = 1;
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, 2);
dec->ackend += 2;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
} else {
#ifdef OLDMST
/*
* MorningStar before v1.3 needs NAK
*/
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->nakend, cp, 2);
dec->nakend += 2;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#else
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->rejend, cp, 2);
dec->rejend += 2;
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->my_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#endif
}
break;
case MODE_NAK:
case MODE_REJ:
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->want_acfcomp = 0;
lcp->his_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
break;
}
break;
case TY_SDP:
LogPrintf(LogLCP, "%s\n", request);
switch (mode_type) {
case MODE_REQ:
case MODE_NAK:
case MODE_REJ:
break;
}
break;
default:
sz = (sizeof desc - 2) / 2;
if (sz > length - 2)
sz = length - 2;
pos = 0;
desc[0] = sz ? ' ' : '\0';
for (pos = 0; sz--; pos++)
sprintf(desc+(pos<<1)+1, "%02x", cp[pos+2]);
LogPrintf(LogLCP, "%s%s\n", request, desc);
if (mode_type == MODE_REQ) {
reqreject:
Remove more globals
1998-03-13 21:08:05 +00:00
if (length > sizeof dec->rej - (dec->rejend - dec->rej)) {
length = sizeof dec->rej - (dec->rejend - dec->rej);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
LogPrintf(LogLCP, "Can't REJ length %d - trunating to %d\n",
cp[1], length);
}
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->rejend, cp, length);
dec->rejend += length;
Isolate fsm callback functions Derive {lcp,ipcp,ccp}state structures from struct fsm where convenient. Create bundle_Down() and remove LcpDown()
1998-02-06 02:24:29 +00:00
lcp->my_reject |= (1 << type);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
if (length != cp[1])
return;
}
break;
}
/* to avoid inf. loop */
if (length == 0) {
LogPrintf(LogLCP, "LCP size zero\n");
break;
}
plen -= length;
cp += length;
}
o Move the accept/deny/disable/enable globals either to the bundle (non-negotiated vars) or to their respective IPCP, LCP or CCP. o Enable rolling throughput statistics by default. o Remove the `display' command. These values now appear in `show bundle', `show ipcp', `show ccp' and `show lcp'. o Initialise auth name & key at bundle create time (oops). o Rename pppd-deflate (the id-24 hack) to deflate24. o Don't send both a REJ and a NAK to an IPCP or LCP REQ. Favour the REJ (already done at the CCP level). o Recurse in datalink_UpdateSet() when we change state, otherwise we end up setting no descriptors and getting jammed in the imminent select() instead of doing the dial/login/hangup. o Display our CHAP encryption method despite being built with DES. o Display VJ as not negotiated in ``show ipcp'' when necessary.
1998-04-16 00:26:21 +00:00
if (mode_type != MODE_NOP)
if (dec->rejend != dec->rej) {
/* rejects are preferred */
dec->ackend = dec->ack;
dec->nakend = dec->nak;
} else if (dec->nakend != dec->nak)
/* then NAKs */
dec->ackend = dec->ack;
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
void
o Don't immediately reply to the first LQR thinking it's a duplicate ('cos it compares with our initial values). o Move the LCP into struct datalink.
1998-03-13 21:07:14 +00:00
LcpInput(struct lcp *lcp, struct mbuf * bp)
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
{
Move the global FSMs into their relevent structures. IPCP, CCP and LCP are now just derived FSMs. Comment each of the FSM implementations so that we can tell what's going on. Revise the state transitions so that CCP and IPCP actually send terminate REQs when appropriate. The OS & IPCP layers are still like spagetti (next job).
1998-01-31 02:48:30 +00:00
/* Got PROTO_LCP from link */
o Don't immediately reply to the first LQR thinking it's a duplicate ('cos it compares with our initial values). o Move the LCP into struct datalink.
1998-03-13 21:07:14 +00:00
FsmInput(&lcp->fsm, bp);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
}
Reference in New Issue Copy Permalink