d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
8f95b97072
freebsd-skq/usr.sbin/ppp/lcp.c

1161 lines
34 KiB
C
Raw Normal View History

1995-01-31 06:29:58 +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.
Remove trailing whitespace.
1995-05-30 03:57:47 +00:00
*
$Id$ -> $FreeBSD$
1999-08-28 01:35:59 +00:00
* $FreeBSD$
Remove trailing whitespace.
1995-05-30 03:57:47 +00:00
*
1995-01-31 06:29:58 +00:00
*/
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
Initial RADIUS support (using libradius). See the man page for details. Compiling with -DNORADIUS (the default for `release') removes support. TODO: The functionality in libradius::rad_send_request() needs to be supplied as a set of routines so that ppp doesn't have to wait indefinitely for the radius server(s). Instead, we need to get a descriptor back, select() on the descriptor, and ask libradius to service it when necessary. For now, ppp blocks SIGALRM while in rad_send_request(), so it misses PAP/CHAP retries & timeouts if they occur. Only PAP is functional. When CHAP is attempted, libradius complains that no User-Password has been specified... rfc2138 says that it *mustn't* be used for CHAP :-( Sponsored by: Internet Business Solutions Ltd., Switzerland
1999-01-28 01:56:34 +00:00
#include <sys/param.h>
Remove unused #includes. Make various bits static. Remove unused variables. Submitted by: eivind
1998-01-21 02:15:33 +00:00
#include <netinet/in.h>
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
#include <netinet/in_systm.h>
#include <netinet/ip.h>
o Add the link name to modem diagnostics. o Create struct mpserver as part of struct mp. mpserver creates a unix-domain socket based on the peers auth name and endpoint discriminator. If it already exists, ppp will ``pass the link'' over to the owner of the socket, joining it into the bundle of another ppp invocation, otherwise ppp waits for other invocations to pass it links through this socket. The final piece of code will be the code that flattens our datalink info and passes it down this channel (not yet implemented).
1998-04-28 01:25:46 +00:00
#include <sys/un.h>
Cosmetic (no functional changes): o Add missing $Id$s o Move extern decls from .c -> .h files o Staticize o Remove #includes from .h files o style(9)ify includes o bcopy -> memcpy bzero -> memset bcmp -> memcmp index -> strchr rindex -> strrchr o Move timeout.h -> timer.h (making it consistent w/ timer.c) o Add -Wmissing-prototypes
1997-10-26 01:04:02 +00:00
Don't pause for one second before starting LCP negotiation. Instead, incrementally pause after receiving LCPs with the same magic. We can now suffer a server that waits more than 1 second before responding. Pauses greater than a second get hopelessly confusing as when the server eventually starts, it sees a flood of Config Requests followed by config NAKs and changes of magic. This causes the server to change its magic over and over.....
1997-09-22 02:51:24 +00:00
#include <signal.h>
Cosmetic (no functional changes): o Add missing $Id$s o Move extern decls from .c -> .h files o Staticize o Remove #includes from .h files o style(9)ify includes o bcopy -> memcpy bzero -> memset bcmp -> memcmp index -> strchr rindex -> strrchr o Move timeout.h -> timer.h (making it consistent w/ timer.c) o Add -Wmissing-prototypes
1997-10-26 01:04:02 +00:00
#include <stdio.h>
#include <stdlib.h>
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
#include <string.h>
Cosmetic (no functional changes): o Add missing $Id$s o Move extern decls from .c -> .h files o Staticize o Remove #includes from .h files o style(9)ify includes o bcopy -> memcpy bzero -> memset bcmp -> memcmp index -> strchr rindex -> strrchr o Move timeout.h -> timer.h (making it consistent w/ timer.c) o Add -Wmissing-prototypes
1997-10-26 01:04:02 +00:00
#include <termios.h>
#include <unistd.h>
o Redesign the layering mechanism and make the aliasing code part of the layering. We now ``stack'' layers as soon as we open the device (when we figure out what we're dealing with). A static set of `dispatch' routines are also declared for dealing with incoming packets after they've been `pulled' up through the stacked layers. Physical devices are now assigned handlers based on the device type when they're opened. For the moment there are three device types; ttys, execs and tcps. o Increment version number to 2.2 o Make an entry in [uw]tmp for non-tty -direct invocations (after pap/chap authentication). o Make throughput counters quad_t's o Account for the absolute number of mbuf malloc()s and free()s in ``show mem''. o ``show modem'' becomes ``show physical''.
1999-05-08 11:07:56 +00:00
#include "layer.h"
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
#include "ua.h"
Fix a rather nasty use of `static'. This caused a SEGV when running ``link * load label'' as we ended up recursing back into command_Interpret after nuking our command arg list.
1998-06-15 19:06:25 +00:00
#include "defs.h"
Fix prototypes. Remove extraneous decls. Add ``const'' to several places. Allow ``make NOALIAS=1'' to remove IP aliasing. Merge with OpenBSD - only the Makefiles vary. We can now survive a compile with -Wall -Wbad-function-cast -Wcast-align -Wcast-qual -Winline -Wmissing-declarations -Wmissing-prototypes -Wnested-externs -Wpointer-arith -Wredundant-decls -Wshadow -Wstrict-prototypes -Wwrite-strings -Wchar-subscripts (although the Makefile just contains -Wall).
1997-11-22 03:37:54 +00:00
#include "command.h"
Cosmetic (no functional changes): o Add missing $Id$s o Move extern decls from .c -> .h files o Staticize o Remove #includes from .h files o style(9)ify includes o bcopy -> memcpy bzero -> memset bcmp -> memcmp index -> strchr rindex -> strrchr o Move timeout.h -> timer.h (making it consistent w/ timer.c) o Add -Wmissing-prototypes
1997-10-26 01:04:02 +00:00
#include "mbuf.h"
#include "log.h"
#include "timer.h"
1995-01-31 06:29:58 +00:00
#include "fsm.h"
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
#include "iplist.h"
1995-01-31 06:29:58 +00:00
#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"
o Redesign the layering mechanism and make the aliasing code part of the layering. We now ``stack'' layers as soon as we open the device (when we figure out what we're dealing with). A static set of `dispatch' routines are also declared for dealing with incoming packets after they've been `pulled' up through the stacked layers. Physical devices are now assigned handlers based on the device type when they're opened. For the moment there are three device types; ttys, execs and tcps. o Increment version number to 2.2 o Make an entry in [uw]tmp for non-tty -direct invocations (after pap/chap authentication). o Make throughput counters quad_t's o Account for the absolute number of mbuf malloc()s and free()s in ``show mem''. o ``show modem'' becomes ``show physical''.
1999-05-08 11:07:56 +00:00
#include "proto.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"
1995-01-31 06:29:58 +00:00
#include "hdlc.h"
Final cleanup for now. -Wall is now silent. A couple of bogons found.
1996-01-11 17:48:59 +00:00
#include "ccp.h"
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
#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"
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
#include "slcompress.h"
#include "ipcp.h"
#include "filter.h"
#include "mp.h"
o Defer setting up pap/chap based IP numbers & labels until after we've determined if we're going to join another ppp invocation. o Make ``show link'' show all link details, and ``show links'' just give a list of links and their current status. o Show our current label in ``show bundle''. o Allow link cloning and removal as soon as our MRRU is set. o Make ``show lcp'' require context as nothing will ever change in our MP LCP (it's auto-configured as per rfc1990). o Initialise our LQM owner in hdlc_Init(). o Store our endpoint discriminator and authentication name at both the datalink and multilink level and compare them when we've finished AUTHENTICATE phase and before entering NETWORK phase. If they don't match, close the link again. Display the information in the appropriate ``show'' command. o Initialise datalink::phone and datalink::fsmp.object properly when we're cloning the link. o Show which link we're passing LQRs on in our diagnostics. o Reject endpoint discriminator REQs at the logical multilink level. o Remove the rest of our CARRIER and LINK logging setup.
1998-04-24 19:16:15 +00:00
#include "chat.h"
New user Process PPP based on iij-ppp0.94beta2. o Supporting SYNC SIO device (But need a device driver) - add "set speed sync" o Fixing bug for Predictor-1 function. o Add new parameter that re-sent interval for set timeout commands. o Improving RTT (Round Trip Time) and reducing processor time. - Previous Timer service was using polling, and now using SIGALRM ;-) - A 0.94beta2 will not work correctly.... -- Follows are additinal feature not including 0.94beta2 o Support Proxy ARP - add "enable/disable proxy" commands o Marging common routine in CHAP/PAP. o Enhancing LCP/IPCP log information. o Support local Authfication connection on port 300x and tty. - You can set up pair of your "hostname -s" and password in ppp.secret. if either ppp.secret file nor your hostname line don't exist, It will notify a message and working as same as previous version.(Backword compatibility) - If you did set up them, It's allow connection but nothing to do except help and passwd command. - add "passwd yourpasswd" commands o Support afilter - keep Alive filter that a packet can send/receiving according to ifilter/ofilter but doesn't count it as preventing idle timer expires. - Same syntax of other filters. o Fixing bugs reported by current user for previous one. Thanks !! Reviewed by: Atsushi Murai (amurai@spec.co.jp)
1995-02-26 12:18:08 +00:00
#include "auth.h"
Cosmetic (no functional changes): o Add missing $Id$s o Move extern decls from .c -> .h files o Staticize o Remove #includes from .h files o style(9)ify includes o bcopy -> memcpy bzero -> memset bcmp -> memcmp index -> strchr rindex -> strrchr o Move timeout.h -> timer.h (making it consistent w/ timer.c) o Add -Wmissing-prototypes
1997-10-26 01:04:02 +00:00
#include "chap.h"
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
#include "cbcp.h"
o Defer setting up pap/chap based IP numbers & labels until after we've determined if we're going to join another ppp invocation. o Make ``show link'' show all link details, and ``show links'' just give a list of links and their current status. o Show our current label in ``show bundle''. o Allow link cloning and removal as soon as our MRRU is set. o Make ``show lcp'' require context as nothing will ever change in our MP LCP (it's auto-configured as per rfc1990). o Initialise our LQM owner in hdlc_Init(). o Store our endpoint discriminator and authentication name at both the datalink and multilink level and compare them when we've finished AUTHENTICATE phase and before entering NETWORK phase. If they don't match, close the link again. Display the information in the appropriate ``show'' command. o Initialise datalink::phone and datalink::fsmp.object properly when we're cloning the link. o Show which link we're passing LQRs on in our diagnostics. o Reject endpoint discriminator REQs at the logical multilink level. o Remove the rest of our CARRIER and LINK logging setup.
1998-04-24 19:16:15 +00:00
#include "datalink.h"
Initial RADIUS support (using libradius). See the man page for details. Compiling with -DNORADIUS (the default for `release') removes support. TODO: The functionality in libradius::rad_send_request() needs to be supplied as a set of routines so that ppp doesn't have to wait indefinitely for the radius server(s). Instead, we need to get a descriptor back, select() on the descriptor, and ask libradius to service it when necessary. For now, ppp blocks SIGALRM while in rad_send_request(), so it misses PAP/CHAP retries & timeouts if they occur. Only PAP is functional. When CHAP is attempted, libradius complains that no User-Password has been specified... rfc2138 says that it *mustn't* be used for CHAP :-( Sponsored by: Internet Business Solutions Ltd., Switzerland
1999-01-28 01:56:34 +00:00
#ifndef NORADIUS
#include "radius.h"
#endif
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
#include "bundle.h"
1995-01-31 06:29:58 +00:00
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
/* for received LQRs */
struct lqrreq {
u_char type;
u_char length;
u_short proto; /* Quality protocol */
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
u_int32_t period; /* Reporting interval */
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
};
o Create a new ``timer'' log level. This lets us ``set log debug'' without filling our filesystem/screen with junk that we don't really want to see. o change PHYS_STDIN to PHYS_DIRECT - we can handle incoming connections that aren't on STDIN_FILENO now. o Allow return values from our FSM LayerUp functions. If LayerUp() fails, the FSM does an immediate FsmDown() without calling the fsm_parent's Layer{Up,Down} functions. o Clear the close-on-exec flag of file descriptor 3 when executing chat programs so that our documented ability to communicate with /dev/tty via that descriptor works. Also document it as descriptor 3, not 4 :-O o Allow a ``rm'' command as an alias for ``remove''. o Fix the bind()/connect()/accept() calls made by the MP server. o Create bundle_SendDatalink() and bundle_ReceiveDatalink(). This allows `struct datalink's to flatten themselves, pass through a pipe (read: the eye of a needle !) and come alive at the other end. The donator then fork()s & exec()s pppmpipe, ``passing'' the connection to another ppp instance. *** PPP NOW TALKS MULTILINK :-))) *** Our link utilization is hideous, and lots of code needs tidying still. It's also probably riddled with bugs ! It's been tested against itself only, and has hung once, so confidence isn't high....
1998-04-30 23:53:56 +00:00
static int LcpLayerUp(struct fsm *);
Overhaul ppp: o Use syslog o Remove references to stdout/stderr (incl perror()) o Introduce VarTerm - the interactive terminal or zero o Allow "set timeout" to affect current session o Change "set debug" to "set log" o Allow "set log [+|-]flag" o Make MSEXT and PASSWDAUTH stuff the default o Move all #ifdef DEBUG stuff into the code - this shouldn't be too much overhead. It's now controlled with "set log +debug" o Add "set log command, debug, tun, warn, error, alert" o Remove cdefs.h, and assume an ansi compiler. o Improve all diagnostic output o Don't trap SIGSEGV o SIGHUP now terminates again (log files are controlled by syslog) o Call CloseModem() when changing devices o Fix parsing of third arg of "delete" I think this fixes the "magic is same" problems that some people have been experiencing. The man page is being rewritten. It'll follow soon.
1997-06-09 03:27:43 +00:00
static void LcpLayerDown(struct fsm *);
static void LcpLayerStart(struct fsm *);
static void LcpLayerFinish(struct fsm *);
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
static void LcpInitRestartCounter(struct fsm *, int);
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 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,
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
fsm_NullRecvResetReq,
fsm_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
o Create a new ``timer'' log level. This lets us ``set log debug'' without filling our filesystem/screen with junk that we don't really want to see. o change PHYS_STDIN to PHYS_DIRECT - we can handle incoming connections that aren't on STDIN_FILENO now. o Allow return values from our FSM LayerUp functions. If LayerUp() fails, the FSM does an immediate FsmDown() without calling the fsm_parent's Layer{Up,Down} functions. o Clear the close-on-exec flag of file descriptor 3 when executing chat programs so that our documented ability to communicate with /dev/tty via that descriptor works. Also document it as descriptor 3, not 4 :-O o Allow a ``rm'' command as an alias for ``remove''. o Fix the bind()/connect()/accept() calls made by the MP server. o Create bundle_SendDatalink() and bundle_ReceiveDatalink(). This allows `struct datalink's to flatten themselves, pass through a pipe (read: the eye of a needle !) and come alive at the other end. The donator then fork()s & exec()s pppmpipe, ``passing'' the connection to another ppp instance. *** PPP NOW TALKS MULTILINK :-))) *** Our link utilization is hideous, and lots of code needs tidying still. It's also probably riddled with bugs ! It's been tested against itself only, and has hung once, so confidence isn't high....
1998-04-30 23:53:56 +00:00
static const char *lcp_TimerNames[] =
{"LCP restart", "LCP openmode", "LCP stopped"};
1995-01-31 06:29:58 +00:00
Fix prototypes. Remove extraneous decls. Add ``const'' to several places. Allow ``make NOALIAS=1'' to remove IP aliasing. Merge with OpenBSD - only the Makefiles vary. We can now survive a compile with -Wall -Wbad-function-cast -Wcast-align -Wcast-qual -Winline -Wmissing-declarations -Wmissing-prototypes -Wnested-externs -Wpointer-arith -Wredundant-decls -Wshadow -Wstrict-prototypes -Wwrite-strings -Wchar-subscripts (although the Makefile just contains -Wall).
1997-11-22 03:37:54 +00:00
static const char *cftypes[] = {
Update the cftypes arrays according to rfc1700. Log IPCP reqests as IPCP.
1997-11-14 15:39:15 +00:00
/* 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 */
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
"MRRU", /* 17: Multilink-MRRU */
"SHORTSEQ", /* 18: Multilink-Short-Sequence-Number-Header */
Some documentation corrections & typo fixes....
1998-05-29 18:32:41 +00:00
"ENDDISC", /* 19: Multilink-Endpoint-Discriminator */
Update the cftypes arrays according to rfc1700. Log IPCP reqests as IPCP.
1997-11-14 15:39:15 +00:00
"PROPRIETRY", /* 20: Proprietary */
"DCEID", /* 21: DCE-Identifier */
"MULTIPP", /* 22: Multi-Link-Plus-Procedure */
"LDBACP", /* 23: Link Discriminator for BACP */
1995-01-31 06:29:58 +00:00
};
Cosmetic (style): sizeof(var) -> sizeof var sizeof type -> sizeof(type) Suggested by: J Wunsch <j@uriah.heep.sax.de>
1997-12-24 09:29:17 +00:00
#define NCFTYPES (sizeof cftypes/sizeof cftypes[0])
Update the cftypes arrays according to rfc1700. Log IPCP reqests as IPCP.
1997-11-14 15:39:15 +00:00
1995-01-31 06:29:58 +00:00
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)
1995-01-31 06:29:58 +00:00
{
o Add the `set mode' command for change a links current mode. It is not possible to switch to or from dedicated or direct mode, but all other combinations are ok (eg. -auto -> -ddial). o Cope with the fact that commands with optional context may not be able to obtain a link with command_ChooseLink() (if all links have been deleted for example). o Allow `clone'ing in non-multilink mode. We may for example want to configure two links in unilink mode and dial them both, using the one that comes up first. It's also possible to rename ``deflink'' by cloning it, deleting the original, then setting the mode of the new link.
1998-05-15 23:58:30 +00:00
struct link *l;
struct lcp *lcp;
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
l = command_ChooseLink(arg);
o Add the `set mode' command for change a links current mode. It is not possible to switch to or from dedicated or direct mode, but all other combinations are ok (eg. -auto -> -ddial). o Cope with the fact that commands with optional context may not be able to obtain a link with command_ChooseLink() (if all links have been deleted for example). o Allow `clone'ing in non-multilink mode. We may for example want to configure two links in unilink mode and dial them both, using the one that comes up first. It's also possible to rename ``deflink'' by cloning it, deleting the original, then setting the mode of the new link.
1998-05-15 23:58:30 +00:00
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"
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
" MAGIC %08lx, MRRU %u, SHORTSEQ %s, 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",
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
(u_long)lcp->his_magic, lcp->his_mrru,
lcp->his_shortseq ? "on" : "off", 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"
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
" MAGIC %08lx, MRRU %u, SHORTSEQ %s, 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",
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
(u_long)lcp->want_magic, lcp->want_mrru,
lcp->want_shortseq ? "on" : "off", 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);
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);
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
prompt_Printf(arg->prompt, "\n FSM retry = %us, max %u Config"
" REQ%s, %u Term REQ%s\n", lcp->cfg.fsm.timeout,
lcp->cfg.fsm.maxreq, lcp->cfg.fsm.maxreq == 1 ? "" : "s",
lcp->cfg.fsm.maxtrm, lcp->cfg.fsm.maxtrm == 1 ? "" : "s");
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",
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
command_ShowNegval(lcp->cfg.chap05));
#ifdef HAVE_DES
prompt_Printf(arg->prompt, " MSCHAP = %s\n",
command_ShowNegval(lcp->cfg.chap80nt));
prompt_Printf(arg->prompt, " LANMan = %s\n",
command_ShowNegval(lcp->cfg.chap80lm));
#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
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));
Overhaul ppp: o Use syslog o Remove references to stdout/stderr (incl perror()) o Introduce VarTerm - the interactive terminal or zero o Allow "set timeout" to affect current session o Change "set debug" to "set log" o Allow "set log [+|-]flag" o Make MSEXT and PASSWDAUTH stuff the default o Move all #ifdef DEBUG stuff into the code - this shouldn't be too much overhead. It's now controlled with "set log +debug" o Add "set log command, debug, tun, warn, error, alert" o Remove cdefs.h, and assume an ansi compiler. o Improve all diagnostic output o Don't trap SIGSEGV o SIGHUP now terminates again (log files are controlled by syslog) o Call CloseModem() when changing devices o Fix parsing of third arg of "delete" I think this fixes the "magic is same" problems that some people have been experiencing. The man page is being rewritten. It'll follow soon.
1997-06-09 03:27:43 +00:00
return 0;
1995-01-31 06:29:58 +00:00
}
Make things work when sizeof(long) != 32 (hopefully)
1998-01-11 17:50:49 +00:00
static u_int32_t
Fix prototypes. Remove extraneous decls. Add ``const'' to several places. Allow ``make NOALIAS=1'' to remove IP aliasing. Merge with OpenBSD - only the Makefiles vary. We can now survive a compile with -Wall -Wbad-function-cast -Wcast-align -Wcast-qual -Winline -Wmissing-declarations -Wmissing-prototypes -Wnested-externs -Wpointer-arith -Wredundant-decls -Wshadow -Wstrict-prototypes -Wwrite-strings -Wchar-subscripts (although the Makefile just contains -Wall).
1997-11-22 03:37:54 +00:00
GenerateMagic(void)
1995-01-31 06:29:58 +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
/* Generate random number which will be used as magic number */
Cosmetic (no functional changes): o Add missing $Id$s o Move extern decls from .c -> .h files o Staticize o Remove #includes from .h files o style(9)ify includes o bcopy -> memcpy bzero -> memset bcmp -> memcmp index -> strchr rindex -> strrchr o Move timeout.h -> timer.h (making it consistent w/ timer.c) o Add -Wmissing-prototypes
1997-10-26 01:04:02 +00:00
randinit();
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
return random();
1995-01-31 06:29:58 +00:00
}
void
o Create a new ``timer'' log level. This lets us ``set log debug'' without filling our filesystem/screen with junk that we don't really want to see. o change PHYS_STDIN to PHYS_DIRECT - we can handle incoming connections that aren't on STDIN_FILENO now. o Allow return values from our FSM LayerUp functions. If LayerUp() fails, the FSM does an immediate FsmDown() without calling the fsm_parent's Layer{Up,Down} functions. o Clear the close-on-exec flag of file descriptor 3 when executing chat programs so that our documented ability to communicate with /dev/tty via that descriptor works. Also document it as descriptor 3, not 4 :-O o Allow a ``rm'' command as an alias for ``remove''. o Fix the bind()/connect()/accept() calls made by the MP server. o Create bundle_SendDatalink() and bundle_ReceiveDatalink(). This allows `struct datalink's to flatten themselves, pass through a pipe (read: the eye of a needle !) and come alive at the other end. The donator then fork()s & exec()s pppmpipe, ``passing'' the connection to another ppp instance. *** PPP NOW TALKS MULTILINK :-))) *** Our link utilization is hideous, and lots of code needs tidying still. It's also probably riddled with bugs ! It's been tested against itself only, and has hung once, so confidence isn't high....
1998-04-30 23:53:56 +00:00
lcp_SetupCallbacks(struct lcp *lcp)
1995-01-31 06:29:58 +00:00
{
o Create a new ``timer'' log level. This lets us ``set log debug'' without filling our filesystem/screen with junk that we don't really want to see. o change PHYS_STDIN to PHYS_DIRECT - we can handle incoming connections that aren't on STDIN_FILENO now. o Allow return values from our FSM LayerUp functions. If LayerUp() fails, the FSM does an immediate FsmDown() without calling the fsm_parent's Layer{Up,Down} functions. o Clear the close-on-exec flag of file descriptor 3 when executing chat programs so that our documented ability to communicate with /dev/tty via that descriptor works. Also document it as descriptor 3, not 4 :-O o Allow a ``rm'' command as an alias for ``remove''. o Fix the bind()/connect()/accept() calls made by the MP server. o Create bundle_SendDatalink() and bundle_ReceiveDatalink(). This allows `struct datalink's to flatten themselves, pass through a pipe (read: the eye of a needle !) and come alive at the other end. The donator then fork()s & exec()s pppmpipe, ``passing'' the connection to another ppp instance. *** PPP NOW TALKS MULTILINK :-))) *** Our link utilization is hideous, and lots of code needs tidying still. It's also probably riddled with bugs ! It's been tested against itself only, and has hung once, so confidence isn't high....
1998-04-30 23:53:56 +00:00
lcp->fsm.fn = &lcp_Callbacks;
lcp->fsm.FsmTimer.name = lcp_TimerNames[0];
lcp->fsm.OpenTimer.name = lcp_TimerNames[1];
lcp->fsm.StoppedTimer.name = lcp_TimerNames[2];
}
1995-01-31 06:29:58 +00:00
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
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;
1995-01-31 06:29:58 +00:00
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
fsm_Init(&lcp->fsm, "LCP", PROTO_LCP, mincode, LCP_MAXCODE, LogLCP,
o Create a new ``timer'' log level. This lets us ``set log debug'' without filling our filesystem/screen with junk that we don't really want to see. o change PHYS_STDIN to PHYS_DIRECT - we can handle incoming connections that aren't on STDIN_FILENO now. o Allow return values from our FSM LayerUp functions. If LayerUp() fails, the FSM does an immediate FsmDown() without calling the fsm_parent's Layer{Up,Down} functions. o Clear the close-on-exec flag of file descriptor 3 when executing chat programs so that our documented ability to communicate with /dev/tty via that descriptor works. Also document it as descriptor 3, not 4 :-O o Allow a ``rm'' command as an alias for ``remove''. o Fix the bind()/connect()/accept() calls made by the MP server. o Create bundle_SendDatalink() and bundle_ReceiveDatalink(). This allows `struct datalink's to flatten themselves, pass through a pipe (read: the eye of a needle !) and come alive at the other end. The donator then fork()s & exec()s pppmpipe, ``passing'' the connection to another ppp instance. *** PPP NOW TALKS MULTILINK :-))) *** Our link utilization is hideous, and lots of code needs tidying still. It's also probably riddled with bugs ! It's been tested against itself only, and has hung once, so confidence isn't high....
1998-04-30 23:53:56 +00:00
bundle, l, parent, &lcp_Callbacks, lcp_TimerNames);
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.accmap = 0;
lcp->cfg.openmode = 1;
lcp->cfg.lqrperiod = DEF_LQRPERIOD;
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
lcp->cfg.fsm.timeout = DEF_FSMRETRY;
lcp->cfg.fsm.maxreq = DEF_FSMTRIES;
lcp->cfg.fsm.maxtrm = DEF_FSMTRIES;
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
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;
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
lcp->cfg.chap05 = NEG_ACCEPTED;
#ifdef HAVE_DES
lcp->cfg.chap80nt = NEG_ACCEPTED;
lcp->cfg.chap80lm = NEG_ACCEPTED;
#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
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);
1995-01-31 06:29:58 +00:00
}
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)
1995-01-31 06:29:58 +00:00
{
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->fsm.open_mode = openmode;
Initialise lcp::his_mru to the ``set mtu'' value if it's less than DEF_MRU, allowing our interface mtu to be decreased despite negotiation with the peer.
1998-09-09 00:03:09 +00:00
lcp->his_mru = lcp->fsm.bundle->cfg.mtu;
if (!lcp->his_mru || lcp->his_mru > DEF_MRU)
lcp->his_mru = DEF_MRU;
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
lcp->his_mrru = 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_magic = 0;
lcp->his_lqrperiod = 0;
lcp->his_acfcomp = 0;
lcp->his_auth = 0;
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
lcp->his_authtype = 0;
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
lcp->his_callback.opmask = 0;
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
lcp->his_shortseq = 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
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;
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
lcp->want_mrru = lcp->fsm.bundle->ncp.mp.cfg.mrru;
lcp->want_shortseq = IsEnabled(lcp->fsm.bundle->ncp.mp.cfg.shortseq) ? 1 : 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_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) {
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
struct physical *p = link2physical(lcp->fsm.link);
Initialise multilink LCPs according to rfc1990.
1998-04-03 19:26:12 +00:00
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();
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
if (IsEnabled(lcp->cfg.chap05)) {
lcp->want_auth = PROTO_CHAP;
lcp->want_authtype = 0x05;
#ifdef HAVE_DES
} else if (IsEnabled(lcp->cfg.chap80nt) ||
IsEnabled(lcp->cfg.chap80lm)) {
lcp->want_auth = PROTO_CHAP;
lcp->want_authtype = 0x80;
#endif
} else if (IsEnabled(lcp->cfg.pap)) {
lcp->want_auth = PROTO_PAP;
lcp->want_authtype = 0;
} else {
lcp->want_auth = 0;
lcp->want_authtype = 0;
}
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
if (p->type != PHYS_DIRECT)
memcpy(&lcp->want_callback, &p->dl->cfg.callback, sizeof(struct callback));
else
lcp->want_callback.opmask = 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_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;
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
lcp->want_authtype = 0;
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
lcp->want_callback.opmask = 0;
Initialise multilink LCPs according to rfc1990.
1998-04-03 19:26:12 +00:00
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;
1995-01-31 06:29:58 +00:00
}
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
static void
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
LcpInitRestartCounter(struct fsm *fp, int what)
1995-01-31 06:29:58 +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
/* 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);
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
fp->FsmTimer.load = lcp->cfg.fsm.timeout * SECTICKS;
switch (what) {
case FSM_REQ_TIMER:
fp->restart = lcp->cfg.fsm.maxreq;
break;
case FSM_TRM_TIMER:
fp->restart = lcp->cfg.fsm.maxtrm;
break;
default:
fp->restart = 1;
break;
}
1995-01-31 06:29:58 +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
LcpSendConfigReq(struct fsm *fp)
1995-01-31 06:29:58 +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);
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
u_char buff[200];
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
struct lcp_opt *o;
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
struct mp *mp;
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
u_int16_t proto;
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) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogERROR, "%s: LcpSendConfigReq: Not a physical link !\n",
o Remove LINK level logging. o change the default link name to ``deflink'' rather than ``default''. o Prepend the link name to CCP and LCP FSM diagnostics. o Protect against 0 length options in CCP and IPCP REQ interpreters (already done for LCP). o Allow optional context for the `show' command. o Use MPs link when interpreting commands if the multilink mrru is configured rather than when multilink is active. This means that once we've ``set mrru xxx'', we then need to ``link deflink show ccp'' etc if we want to do link-level stuff (based on the command requiring optional or manditory context). o Use the ifconfig'd interface address in `set enddisc {ip,mac}' if it's there, otherwise the configuration file value.
1998-04-24 19:15:48 +00:00
fp->link->name);
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
return;
}
1995-01-31 06:29:58 +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
o = (struct lcp_opt *)buff;
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
if (!physical_IsSync(p)) {
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +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);
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +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);
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +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)) {
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
ua_htonl(&lcp->want_accmap, o->data);
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_ACCMAP, 6, o);
}
1995-01-31 06:29:58 +00:00
}
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +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)) {
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
ua_htons(&lcp->want_mru, o->data);
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_MRU, 4, o);
}
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +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)) {
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
ua_htonl(&lcp->want_magic, o->data);
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_MAGICNUM, 6, o);
}
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +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)) {
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
proto = PROTO_LQR;
ua_htons(&proto, o->data);
ua_htonl(&lcp->want_lqrperiod, o->data + 2);
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_QUALPROTO, 8, o);
}
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
switch (lcp->want_auth) {
case PROTO_PAP:
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
proto = PROTO_PAP;
ua_htons(&proto, o->data);
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, 4, o);
1995-01-31 06:29:58 +00:00
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
case PROTO_CHAP:
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
proto = PROTO_CHAP;
ua_htons(&proto, o->data);
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
o->data[2] = lcp->want_authtype;
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);
1995-01-31 06:29:58 +00:00
break;
}
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
if (!REJECTED(lcp, TY_CALLBACK)) {
if (lcp->want_callback.opmask & CALLBACK_BIT(CALLBACK_AUTH)) {
*o->data = CALLBACK_AUTH;
INC_LCP_OPT(TY_CALLBACK, 3, o);
} else if (lcp->want_callback.opmask & CALLBACK_BIT(CALLBACK_CBCP)) {
*o->data = CALLBACK_CBCP;
INC_LCP_OPT(TY_CALLBACK, 3, o);
} else if (lcp->want_callback.opmask & CALLBACK_BIT(CALLBACK_E164)) {
int sz = strlen(lcp->want_callback.msg);
if (sz > sizeof o->data - 1) {
sz = sizeof o->data - 1;
log_Printf(LogWARN, "Truncating E164 data to %d octets (oops!)\n", sz);
}
*o->data = CALLBACK_E164;
memcpy(o->data + 1, lcp->want_callback.msg, sz);
INC_LCP_OPT(TY_CALLBACK, sz + 3, o);
}
}
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
if (lcp->want_mrru && !REJECTED(lcp, TY_MRRU)) {
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
ua_htons(&lcp->want_mrru, o->data);
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
INC_LCP_OPT(TY_MRRU, 4, o);
1995-01-31 06:29:58 +00:00
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
if (lcp->want_shortseq && !REJECTED(lcp, TY_SHORTSEQ))
INC_LCP_OPT(TY_SHORTSEQ, 2, o);
}
1995-01-31 06:29:58 +00:00
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
mp = &lcp->fsm.bundle->ncp.mp;
Allow our endpoint discriminator to be enabled, disabled, accepted and denied. This is necessary for some MP implementations that get confused if you accept their endpoint discriminator but reject their MRRU.
1999-06-09 16:54:04 +00:00
if (mp->cfg.enddisc.class != 0 && IsEnabled(mp->cfg.negenddisc) &&
!REJECTED(lcp, TY_ENDDISC)) {
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
*o->data = mp->cfg.enddisc.class;
memcpy(o->data+1, mp->cfg.enddisc.address, mp->cfg.enddisc.len);
INC_LCP_OPT(TY_ENDDISC, mp->cfg.enddisc.len + 3, o);
}
1995-01-31 06:29:58 +00:00
o Alter the mbuf type as it's processed by different layers. o Show more information about missing MP fragments in ``show mp''. o Do away with mbuf_Log(). It was showing mbuf stats twice on receipt of LCP/CCP/IPCP packets.... ???!!? o Pre-allocate a bit extra when creating LQR packets to avoid having to allocate another mbuf in mbuf_Prepend().
1999-06-02 15:59:09 +00:00
fsm_Output(fp, CODE_CONFIGREQ, fp->reqid, buff, (u_char *)o - buff,
MB_LCPOUT);
1995-01-31 06:29:58 +00:00
}
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)
1995-01-31 06:29:58 +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 Alter the mbuf type as it's processed by different layers. o Show more information about missing MP fragments in ``show mp''. o Do away with mbuf_Log(). It was showing mbuf stats twice on receipt of LCP/CCP/IPCP packets.... ???!!? o Pre-allocate a bit extra when creating LQR packets to avoid having to allocate another mbuf in mbuf_Prepend().
1999-06-02 15:59:09 +00:00
fsm_Output(&lcp->fsm, CODE_PROTOREJ, lcp->fsm.reqid, option, count,
MB_LCPOUT);
1995-01-31 06:29:58 +00:00
}
New user Process PPP based on iij-ppp0.94beta2. o Supporting SYNC SIO device (But need a device driver) - add "set speed sync" o Fixing bug for Predictor-1 function. o Add new parameter that re-sent interval for set timeout commands. o Improving RTT (Round Trip Time) and reducing processor time. - Previous Timer service was using polling, and now using SIGALRM ;-) - A 0.94beta2 will not work correctly.... -- Follows are additinal feature not including 0.94beta2 o Support Proxy ARP - add "enable/disable proxy" commands o Marging common routine in CHAP/PAP. o Enhancing LCP/IPCP log information. o Support local Authfication connection on port 300x and tty. - You can set up pair of your "hostname -s" and password in ppp.secret. if either ppp.secret file nor your hostname line don't exist, It will notify a message and working as same as previous version.(Backword compatibility) - If you did set up them, It's allow connection but nothing to do except help and passwd command. - add "passwd yourpasswd" commands o Support afilter - keep Alive filter that a packet can send/receiving according to ifilter/ofilter but doesn't count it as preventing idle timer expires. - Same syntax of other filters. o Fixing bugs reported by current user for previous one. Thanks !! Reviewed by: Atsushi Murai (amurai@spec.co.jp)
1995-02-26 12:18:08 +00:00
static void
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
LcpSentTerminateReq(struct fsm *fp)
New user Process PPP based on iij-ppp0.94beta2. o Supporting SYNC SIO device (But need a device driver) - add "set speed sync" o Fixing bug for Predictor-1 function. o Add new parameter that re-sent interval for set timeout commands. o Improving RTT (Round Trip Time) and reducing processor time. - Previous Timer service was using polling, and now using SIGALRM ;-) - A 0.94beta2 will not work correctly.... -- Follows are additinal feature not including 0.94beta2 o Support Proxy ARP - add "enable/disable proxy" commands o Marging common routine in CHAP/PAP. o Enhancing LCP/IPCP log information. o Support local Authfication connection on port 300x and tty. - You can set up pair of your "hostname -s" and password in ppp.secret. if either ppp.secret file nor your hostname line don't exist, It will notify a message and working as same as previous version.(Backword compatibility) - If you did set up them, It's allow connection but nothing to do except help and passwd command. - add "passwd yourpasswd" commands o Support afilter - keep Alive filter that a packet can send/receiving according to ifilter/ofilter but doesn't count it as preventing idle timer expires. - Same syntax of other filters. o Fixing bugs reported by current user for previous one. Thanks !! Reviewed by: Atsushi Murai (amurai@spec.co.jp)
1995-02-26 12:18:08 +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 */
New user Process PPP based on iij-ppp0.94beta2. o Supporting SYNC SIO device (But need a device driver) - add "set speed sync" o Fixing bug for Predictor-1 function. o Add new parameter that re-sent interval for set timeout commands. o Improving RTT (Round Trip Time) and reducing processor time. - Previous Timer service was using polling, and now using SIGALRM ;-) - A 0.94beta2 will not work correctly.... -- Follows are additinal feature not including 0.94beta2 o Support Proxy ARP - add "enable/disable proxy" commands o Marging common routine in CHAP/PAP. o Enhancing LCP/IPCP log information. o Support local Authfication connection on port 300x and tty. - You can set up pair of your "hostname -s" and password in ppp.secret. if either ppp.secret file nor your hostname line don't exist, It will notify a message and working as same as previous version.(Backword compatibility) - If you did set up them, It's allow connection but nothing to do except help and passwd command. - add "passwd yourpasswd" commands o Support afilter - keep Alive filter that a packet can send/receiving according to ifilter/ofilter but doesn't count it as preventing idle timer expires. - Same syntax of other filters. o Fixing bugs reported by current user for previous one. Thanks !! Reviewed by: Atsushi Murai (amurai@spec.co.jp)
1995-02-26 12:18:08 +00:00
}
1995-01-31 06:29:58 +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)
1995-01-31 06:29:58 +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 Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
struct physical *p = link2physical(fp->link);
if (p && p->dl->state == DATALINK_CBCP)
cbcp_ReceiveTerminateReq(p);
o Alter the mbuf type as it's processed by different layers. o Show more information about missing MP fragments in ``show mp''. o Do away with mbuf_Log(). It was showing mbuf stats twice on receipt of LCP/CCP/IPCP packets.... ???!!? o Pre-allocate a bit extra when creating LQR packets to avoid having to allocate another mbuf in mbuf_Prepend().
1999-06-02 15:59:09 +00:00
fsm_Output(fp, CODE_TERMACK, id, NULL, 0, MB_LCPOUT);
1995-01-31 06:29:58 +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)
1995-01-31 06:29:58 +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);
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
log_Printf(LogLCP, "%s: LayerStart\n", fp->link->name);
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;
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
fp->more.reqs = fp->more.naks = fp->more.rejs = lcp->cfg.fsm.maxreq * 3;
1995-01-31 06:29:58 +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)
1995-01-31 06:29:58 +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 */
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
log_Printf(LogLCP, "%s: LayerFinish\n", fp->link->name);
1995-01-31 06:29:58 +00:00
}
o Create a new ``timer'' log level. This lets us ``set log debug'' without filling our filesystem/screen with junk that we don't really want to see. o change PHYS_STDIN to PHYS_DIRECT - we can handle incoming connections that aren't on STDIN_FILENO now. o Allow return values from our FSM LayerUp functions. If LayerUp() fails, the FSM does an immediate FsmDown() without calling the fsm_parent's Layer{Up,Down} functions. o Clear the close-on-exec flag of file descriptor 3 when executing chat programs so that our documented ability to communicate with /dev/tty via that descriptor works. Also document it as descriptor 3, not 4 :-O o Allow a ``rm'' command as an alias for ``remove''. o Fix the bind()/connect()/accept() calls made by the MP server. o Create bundle_SendDatalink() and bundle_ReceiveDatalink(). This allows `struct datalink's to flatten themselves, pass through a pipe (read: the eye of a needle !) and come alive at the other end. The donator then fork()s & exec()s pppmpipe, ``passing'' the connection to another ppp instance. *** PPP NOW TALKS MULTILINK :-))) *** Our link utilization is hideous, and lots of code needs tidying still. It's also probably riddled with bugs ! It's been tested against itself only, and has hung once, so confidence isn't high....
1998-04-30 23:53:56 +00:00
static int
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)
1995-01-31 06:29:58 +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
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
log_Printf(LogLCP, "%s: LayerUp\n", fp->link->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
async_SetLinkParams(&p->async, lcp);
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
lqr_Start(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
hdlc_StartTimer(&p->hdlc);
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
fp->more.reqs = fp->more.naks = fp->more.rejs = lcp->cfg.fsm.maxreq * 3;
o Create a new ``timer'' log level. This lets us ``set log debug'' without filling our filesystem/screen with junk that we don't really want to see. o change PHYS_STDIN to PHYS_DIRECT - we can handle incoming connections that aren't on STDIN_FILENO now. o Allow return values from our FSM LayerUp functions. If LayerUp() fails, the FSM does an immediate FsmDown() without calling the fsm_parent's Layer{Up,Down} functions. o Clear the close-on-exec flag of file descriptor 3 when executing chat programs so that our documented ability to communicate with /dev/tty via that descriptor works. Also document it as descriptor 3, not 4 :-O o Allow a ``rm'' command as an alias for ``remove''. o Fix the bind()/connect()/accept() calls made by the MP server. o Create bundle_SendDatalink() and bundle_ReceiveDatalink(). This allows `struct datalink's to flatten themselves, pass through a pipe (read: the eye of a needle !) and come alive at the other end. The donator then fork()s & exec()s pppmpipe, ``passing'' the connection to another ppp instance. *** PPP NOW TALKS MULTILINK :-))) *** Our link utilization is hideous, and lots of code needs tidying still. It's also probably riddled with bugs ! It's been tested against itself only, and has hung once, so confidence isn't high....
1998-04-30 23:53:56 +00:00
return 1;
1995-01-31 06:29:58 +00:00
}
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)
1995-01-31 06:29:58 +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);
1995-01-31 06:29:58 +00:00
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
log_Printf(LogLCP, "%s: LayerDown\n", fp->link->name);
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);
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
lqr_StopTimer(p);
Add ``ipcp'' as an optional argument to ``open'', and make open capable of re-negotiatiating the various layers. It is now possible to change various link options and then re-open the relevant layer, making the changes effective - for example, switching off VJ compression or starting ECHO LQRs on-the-fly.
1998-06-25 22:33:31 +00:00
lcp_Setup(fsm2lcp(fp), 0);
1995-01-31 06:29:58 +00:00
}
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
static int
E164ok(struct callback *cb, char *req, int sz)
{
char list[sizeof cb->msg], *next;
int len;
if (!strcmp(cb->msg, "*"))
return 1;
strncpy(list, cb->msg, sizeof list - 1);
list[sizeof list - 1] = '\0';
for (next = strtok(list, ","); next; next = strtok(NULL, ",")) {
len = strlen(next);
if (sz == len && !memcmp(list, req, sz))
return 1;
}
return 0;
}
1995-01-31 06:29:58 +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)
1995-01-31 06:29:58 +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);
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
int type, length, sz, pos, op, callback_req;
o Use __attribute__ (#ifdef __GNUC__) to type-check LogPrintf() and prompt_Printf(), and fix the bits identified as being wrong as a result.
1998-04-19 23:09:03 +00:00
u_int32_t magic, accmap;
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
u_short mtu, mru, proto;
1995-01-31 06:29:58 +00:00
struct lqrreq *req;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
char request[20], desc[22];
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
struct mp *mp;
o Defer setting up pap/chap based IP numbers & labels until after we've determined if we're going to join another ppp invocation. o Make ``show link'' show all link details, and ``show links'' just give a list of links and their current status. o Show our current label in ``show bundle''. o Allow link cloning and removal as soon as our MRRU is set. o Make ``show lcp'' require context as nothing will ever change in our MP LCP (it's auto-configured as per rfc1990). o Initialise our LQM owner in hdlc_Init(). o Store our endpoint discriminator and authentication name at both the datalink and multilink level and compare them when we've finished AUTHENTICATE phase and before entering NETWORK phase. If they don't match, close the link again. Display the information in the appropriate ``show'' command. o Initialise datalink::phone and datalink::fsmp.object properly when we're cloning the link. o Show which link we're passing LQRs on in our diagnostics. o Reject endpoint discriminator REQs at the logical multilink level. o Remove the rest of our CARRIER and LINK logging setup.
1998-04-24 19:16:15 +00:00
struct physical *p = link2physical(fp->link);
1995-01-31 06:29:58 +00:00
Silence a warning about uninitialized variables.
1999-08-08 17:36:10 +00:00
sz = op = callback_req = 0;
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
1995-01-31 06:29:58 +00:00
while (plen >= sizeof(struct fsmconfig)) {
type = *cp;
length = cp[1];
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
if (type < 0 || type >= NCFTYPES)
snprintf(request, sizeof request, " <%d>[%d]", type, length);
1995-01-31 06:29:58 +00:00
else
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
snprintf(request, sizeof request, " %s[%d]", cftypes[type], length);
1995-01-31 06:29:58 +00:00
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
if (length < 2) {
log_Printf(LogLCP, "%s:%s: Bad LCP length\n", fp->link->name, request);
break;
}
1995-01-31 06:29:58 +00:00
switch (type) {
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
case TY_MRRU:
mp = &lcp->fsm.bundle->ncp.mp;
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
ua_ntohs(cp + 2, &mru);
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s %u\n", request, mru);
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
switch (mode_type) {
case MODE_REQ:
if (mp->cfg.mrru) {
if (REJECTED(lcp, TY_MRRU))
/* Ignore his previous reject so that we REQ next time */
lcp->his_reject &= ~(1 << type);
mtu = lcp->fsm.bundle->cfg.mtu;
if (mru < MIN_MRU || mru < mtu) {
/* Push him up to MTU or MIN_MRU */
lcp->his_mrru = mru < mtu ? mtu : MIN_MRU;
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
memcpy(dec->nakend, cp, 2);
ua_htons(&lcp->his_mrru, dec->nakend + 2);
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
dec->nakend += 4;
} else {
lcp->his_mrru = mtu ? mtu : mru;
memcpy(dec->ackend, cp, 4);
dec->ackend += 4;
}
break;
} else
goto reqreject;
break;
case MODE_NAK:
if (mp->cfg.mrru) {
if (REJECTED(lcp, TY_MRRU))
/* Must have changed his mind ! */
lcp->his_reject &= ~(1 << type);
if (mru > MAX_MRU)
lcp->want_mrru = MAX_MRU;
else if (mru < MIN_MRU)
lcp->want_mrru = MIN_MRU;
else
lcp->want_mrru = mru;
}
/* else we honour our config and don't send the suggested REQ */
break;
case MODE_REJ:
lcp->his_reject |= (1 << type);
o Don't go multilink if the peer hasn't asked for it o Don't go multilink if the peer hasn't agreed o When negotiating multilink, don't try to push the peers MRU up according to our `set mtu' setting. Instead, accept whatever MRU they want and push their MRRU up instead. o When *not* negotiating multilink, REJ short sequence number REQs.
1998-04-23 18:55:50 +00:00
lcp->want_mrru = 0; /* Ah well, no multilink :-( */
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
break;
}
break;
1995-01-31 06:29:58 +00:00
case TY_MRU:
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
ua_ntohs(cp + 2, &mru);
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s %d\n", request, mru);
1995-01-31 06:29:58 +00:00
Don't pass global vars as args. Remove local/global conflicts.
1997-11-08 00:28:11 +00:00
switch (mode_type) {
1995-01-31 06:29:58 +00:00
case MODE_REQ:
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
mtu = lcp->fsm.bundle->cfg.mtu;
o Don't go multilink if the peer hasn't asked for it o Don't go multilink if the peer hasn't agreed o When negotiating multilink, don't try to push the peers MRU up according to our `set mtu' setting. Instead, accept whatever MRU they want and push their MRRU up instead. o When *not* negotiating multilink, REJ short sequence number REQs.
1998-04-23 18:55:50 +00:00
if (mru < MIN_MRU || (!lcp->want_mrru && mru < mtu)) {
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
/* Push him up to MTU or MIN_MRU */
lcp->his_mru = mru < mtu ? mtu : MIN_MRU;
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
memcpy(dec->nakend, cp, 2);
ua_htons(&lcp->his_mru, dec->nakend + 2);
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
dec->nakend += 4;
} else {
lcp->his_mru = mtu ? mtu : mru;
memcpy(dec->ackend, cp, 4);
dec->ackend += 4;
}
1995-01-31 06:29:58 +00:00
break;
case MODE_NAK:
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
if (mru > MAX_MRU)
lcp->want_mru = MAX_MRU;
else if (mru < MIN_MRU)
lcp->want_mru = MIN_MRU;
else
lcp->want_mru = mru;
1995-01-31 06:29:58 +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);
1995-01-31 06:29:58 +00:00
break;
}
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
case TY_ACCMAP:
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
ua_ntohl(cp + 2, &accmap);
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s 0x%08lx\n", request, (u_long)accmap);
1995-01-31 06:29:58 +00:00
Don't pass global vars as args. Remove local/global conflicts.
1997-11-08 00:28:11 +00:00
switch (mode_type) {
1995-01-31 06:29:58 +00:00
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;
Revert the ACCMAP changes where we OR the peers accmap with our own if there are differing bits (last two revisions of lcp.c). This change broke at least one negotiation session. Instead, we just use an OR of the two accmap values when we're doing the ASYNC framing.
1999-04-11 08:51:04 +00:00
memcpy(dec->ackend, cp, 6);
dec->ackend += 6;
1995-01-31 06:29:58 +00:00
break;
case MODE_NAK:
Revert the ACCMAP changes where we OR the peers accmap with our own if there are differing bits (last two revisions of lcp.c). This change broke at least one negotiation session. Instead, we just use an OR of the two accmap values when we're doing the ASYNC framing.
1999-04-11 08:51:04 +00:00
lcp->want_accmap = accmap;
1995-01-31 06:29:58 +00:00
break;
case MODE_REJ:
Revert the ACCMAP changes where we OR the peers accmap with our own if there are differing bits (last two revisions of lcp.c). This change broke at least one negotiation session. Instead, we just use an OR of the two accmap values when we're doing the ASYNC framing.
1999-04-11 08:51:04 +00:00
lcp->his_reject |= (1 << type);
1995-01-31 06:29:58 +00:00
break;
}
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
case TY_AUTHPROTO:
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
ua_ntohs(cp + 2, &proto);
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
log_Printf(LogLCP, "%s 0x%04x (%s)\n", request, proto,
Auth2Nam(proto, length > 4 ? cp[4] : 0));
1995-01-31 06:29:58 +00:00
Don't pass global vars as args. Remove local/global conflicts.
1997-11-08 00:28:11 +00:00
switch (mode_type) {
1995-01-31 06:29:58 +00:00
case MODE_REQ:
switch (proto) {
case PROTO_PAP:
if (length != 4) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, " Bad length!\n");
1995-01-31 06:29:58 +00:00
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;
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
lcp->his_authtype = 0;
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, length);
dec->ackend += length;
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
} else if (IsAccepted(lcp->cfg.chap05)) {
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;
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
#ifdef HAVE_DES
} else if (IsAccepted(lcp->cfg.chap80nt) ||
IsAccepted(lcp->cfg.chap80lm)) {
*dec->nakend++ = *cp;
*dec->nakend++ = 5;
*dec->nakend++ = (unsigned char) (PROTO_CHAP >> 8);
*dec->nakend++ = (unsigned char) PROTO_CHAP;
*dec->nakend++ = 0x80;
#endif
1995-01-31 06:29:58 +00:00
} else
goto reqreject;
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
case PROTO_CHAP:
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
if (length != 5) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, " Bad length!\n");
1995-01-31 06:29:58 +00:00
goto reqreject;
}
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
if ((cp[4] == 0x05 && IsAccepted(lcp->cfg.chap05))
Allow Microsoft CHAP authentication. This is a combination of MD4 & DES. Submitted by: Gabor Kincses <gabor@acm.org>
1997-09-25 00:52:37 +00:00
#ifdef HAVE_DES
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
|| (cp[4] == 0x80 && (IsAccepted(lcp->cfg.chap80nt) ||
(IsAccepted(lcp->cfg.chap80lm))))
Allow Microsoft CHAP authentication. This is a combination of MD4 & DES. Submitted by: Gabor Kincses <gabor@acm.org>
1997-09-25 00:52:37 +00:00
#endif
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +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
lcp->his_auth = proto;
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
lcp->his_authtype = cp[4];
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->ackend, cp, length);
dec->ackend += length;
Warn that the specific chap protocol isn't supported when rejecting CHAP because of byte 5. Suggested by: Daniel O'Callaghan <danny@hilink.com.au>
1998-10-17 12:28:02 +00:00
} else {
#ifndef HAVE_DES
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
if (cp[4] == 0x80)
log_Printf(LogWARN, "CHAP 0x80 not available without DES\n");
else
Warn that the specific chap protocol isn't supported when rejecting CHAP because of byte 5. Suggested by: Daniel O'Callaghan <danny@hilink.com.au>
1998-10-17 12:28:02 +00:00
#endif
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
if (cp[4] != 0x05)
log_Printf(LogWARN, "%s not supported\n",
Auth2Nam(PROTO_CHAP, cp[4]));
if (IsAccepted(lcp->cfg.chap05)) {
*dec->nakend++ = *cp;
*dec->nakend++ = 5;
*dec->nakend++ = (unsigned char) (PROTO_CHAP >> 8);
*dec->nakend++ = (unsigned char) PROTO_CHAP;
*dec->nakend++ = 0x05;
#ifdef HAVE_DES
} else if (IsAccepted(lcp->cfg.chap80nt) ||
IsAccepted(lcp->cfg.chap80lm)) {
*dec->nakend++ = *cp;
*dec->nakend++ = 5;
*dec->nakend++ = (unsigned char) (PROTO_CHAP >> 8);
*dec->nakend++ = (unsigned char) PROTO_CHAP;
*dec->nakend++ = 0x80;
#endif
} else if (IsAccepted(lcp->cfg.pap)) {
Warn that the specific chap protocol isn't supported when rejecting CHAP because of byte 5. Suggested by: Daniel O'Callaghan <danny@hilink.com.au>
1998-10-17 12:28:02 +00:00
*dec->nakend++ = *cp;
*dec->nakend++ = 4;
*dec->nakend++ = (unsigned char) (PROTO_PAP >> 8);
*dec->nakend++ = (unsigned char) PROTO_PAP;
} else
goto reqreject;
}
1995-01-31 06:29:58 +00:00
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
default:
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s 0x%04x - not recognised, NAK\n",
Fix PAP, CHAP & LQR req (I broke the byte ordering when I did the deflate re-org). Make PAP & CHAP negotiation prettier in the log file. If both PAP & CHAP are `enabled' and the peer NAKs CHAP suggesting PAP, be friendly and REQ PAP the next time. This is in line with the rfc.
1997-12-07 23:55:29 +00:00
request, proto);
Remove more globals
1998-03-13 21:08:05 +00:00
memcpy(dec->nakend, cp, length);
dec->nakend += length;
Make the code format more in line with style(9). Update loadalias to use the new libalias api. Update to version 1.1.
1997-08-25 00:29:32 +00:00
break;
1995-01-31 06:29:58 +00:00
}
break;
case MODE_NAK:
Fix PAP, CHAP & LQR req (I broke the byte ordering when I did the deflate re-org). Make PAP & CHAP negotiation prettier in the log file. If both PAP & CHAP are `enabled' and the peer NAKs CHAP suggesting PAP, be friendly and REQ PAP the next time. This is in line with the rfc.
1997-12-07 23:55:29 +00:00
switch (proto) {
case PROTO_PAP:
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +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;
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
lcp->want_authtype = 0;
} else {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(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);
Fix PAP, CHAP & LQR req (I broke the byte ordering when I did the deflate re-org). Make PAP & CHAP negotiation prettier in the log file. If both PAP & CHAP are `enabled' and the peer NAKs CHAP suggesting PAP, be friendly and REQ PAP the next time. This is in line with the rfc.
1997-12-07 23:55:29 +00:00
}
break;
case PROTO_CHAP:
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
if (cp[4] == 0x05 && IsEnabled(lcp->cfg.chap05)) {
lcp->want_auth = PROTO_CHAP;
lcp->want_authtype = 0x05;
#ifdef HAVE_DES
} else if (cp[4] == 0x80 && (IsEnabled(lcp->cfg.chap80nt) ||
IsEnabled(lcp->cfg.chap80lm))) {
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;
Fully support both NT and LANMan CHAP type 0x80 as both authenticator and authenticatee.
1999-02-18 00:52:15 +00:00
lcp->want_authtype = 0x80;
#endif
} else {
#ifndef HAVE_DES
if (cp[4] == 0x80)
log_Printf(LogLCP, "Peer will only send MSCHAP (not available"
" without DES)\n");
else
#endif
Allow control over the number of ConfigREQ & TermREQ attempts that are made in each of the FSMs (LCP, CCP & IPCP) and the number of REQs/Challenges for PAP/CHAP by accepting more arguments in the ``set {c,ip,l}cpretry'' and ``set {ch,p}apretry'' commands. Change the non-convergence thresholds to 3 times the number of configured REQ tries (rather than the previous fixed ``10''). We now notice repeated NAKs and REJs rather than just REQs. Don't suggest that CHAP 0x05 isn't supported when it's not configured. Fix some bugs that expose themselves with smaller numbers of retries: o Handle instantaneous disconnects (set device /dev/null) correctly by stopping all fsm timers in fsm2initial. o Don't forget to uu_unlock() devices that are files but are not ttys (set device /dev/zero). Fix a *HORRENDOUS* bug in RFC1661 (already fixed for an Open event in state ``Closed''): According to the state transition table, a RCR+ or RCR- received in the ``Stopped'' state are supposed to InitRestartCounter, SendConfigReq and SendConfig{Ack,Nak}. However, in ``Stopped'', we haven't yet done a TLS (or the last thing we did is a TLF). We must therefore do the TLS at this point ! This was never noticed before because LCP and CCP used not use LayerStart() for anything interesting, and IPCP tends to go into Stopped then get a Down because of an LCP RTR rather than getting a RCR again.
1999-02-26 21:28:14 +00:00
log_Printf(LogLCP, "Peer will only send %s (not %s)\n",
Auth2Nam(PROTO_CHAP, cp[4]),
#ifdef HAVE_DES
cp[4] == 0x80 ? "configured" :
#endif
"supported");
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);
Fix PAP, CHAP & LQR req (I broke the byte ordering when I did the deflate re-org). Make PAP & CHAP negotiation prettier in the log file. If both PAP & CHAP are `enabled' and the peer NAKs CHAP suggesting PAP, be friendly and REQ PAP the next time. This is in line with the rfc.
1997-12-07 23:55:29 +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);
Fix PAP, CHAP & LQR req (I broke the byte ordering when I did the deflate re-org). Make PAP & CHAP negotiation prettier in the log file. If both PAP & CHAP are `enabled' and the peer NAKs CHAP suggesting PAP, be friendly and REQ PAP the next time. This is in line with the rfc.
1997-12-07 23:55:29 +00:00
break;
}
1995-01-31 06:29:58 +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);
1995-01-31 06:29:58 +00:00
break;
}
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
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;
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
log_Printf(LogLCP, "%s proto %x, interval %lums\n",
request, ntohs(req->proto), (u_long)ntohl(req->period) * 10);
Don't pass global vars as args. Remove local/global conflicts.
1997-11-08 00:28:11 +00:00
switch (mode_type) {
1995-01-31 06:29:58 +00:00
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))
1995-01-31 06:29:58 +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);
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
if (lcp->his_lqrperiod < MIN_LQRPERIOD * 100)
lcp->his_lqrperiod = MIN_LQRPERIOD * 100;
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
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;
1995-01-31 06:29:58 +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);
1995-01-31 06:29:58 +00:00
break;
}
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
case TY_MAGICNUM:
Don't cast potentially unaligned addresses to pointers to non-char types on non-i386 architectures. On Alpha and Sparc we get a bus error if we do.
1998-09-04 18:26:00 +00:00
ua_ntohl(cp + 2, &magic);
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s 0x%08lx\n", request, (u_long)magic);
1995-01-31 06:29:58 +00:00
Don't pass global vars as args. Remove local/global conflicts.
1997-11-08 00:28:11 +00:00
switch (mode_type) {
1995-01-31 06:29:58 +00:00
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) {
Remove initial startup printf's in -direct mode. Make "same magic" error send a NAK (as per rfc).
1997-06-01 03:43:26 +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) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(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);
Don't pause for one second before starting LCP negotiation. Instead, incrementally pause after receiving LCPs with the same magic. We can now suffer a server that waits more than 1 second before responding. Pauses greater than a second get hopelessly confusing as when the server eventually starts, it sees a flood of Config Requests followed by config NAKs and changes of magic. This causes the server to change its magic over and over.....
1997-09-22 02:51:24 +00:00
sigpause(0);
Make the code format more in line with style(9). Update loadalias to use the new libalias api. Update to version 1.1.
1997-08-25 00:29:32 +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_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;
Make the code format more in line with style(9). Update loadalias to use the new libalias api. Update to version 1.1.
1997-08-25 00:29:32 +00:00
}
1995-01-31 06:29:58 +00:00
} else {
goto reqreject;
}
break;
case MODE_NAK:
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(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();
1995-01-31 06:29:58 +00:00
break;
case MODE_REJ:
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(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);
1995-01-31 06:29:58 +00:00
break;
}
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
case TY_PROTOCOMP:
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s\n", request);
1995-01-31 06:29:58 +00:00
Don't pass global vars as args. Remove local/global conflicts.
1997-11-08 00:28:11 +00:00
switch (mode_type) {
1995-01-31 06:29:58 +00:00
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;
Make the code format more in line with style(9). Update loadalias to use the new libalias api. Update to version 1.1.
1997-08-25 00:29:32 +00:00
} else {
1995-01-31 06:29:58 +00:00
#ifdef OLDMST
Deal with the fact that as we now mbuf_Read the fsm header in fsm_Input() we often end up with a NULL mbuf. Deal with a possible NULL mbuf being passed into mbuf_Prepend(). Adjust some spacing to make things more consistent.
1999-05-09 20:02:29 +00:00
/* 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;
1995-01-31 06:29:58 +00:00
#else
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
goto reqreject;
1995-01-31 06:29:58 +00:00
#endif
Make the code format more in line with style(9). Update loadalias to use the new libalias api. Update to version 1.1.
1997-08-25 00:29:32 +00:00
}
1995-01-31 06:29:58 +00:00
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);
1995-01-31 06:29:58 +00:00
break;
}
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
case TY_ACFCOMP:
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s\n", request);
Don't pass global vars as args. Remove local/global conflicts.
1997-11-08 00:28:11 +00:00
switch (mode_type) {
1995-01-31 06:29:58 +00:00
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;
Make the code format more in line with style(9). Update loadalias to use the new libalias api. Update to version 1.1.
1997-08-25 00:29:32 +00:00
} else {
1995-01-31 06:29:58 +00:00
#ifdef OLDMST
Deal with the fact that as we now mbuf_Read the fsm header in fsm_Input() we often end up with a NULL mbuf. Deal with a possible NULL mbuf being passed into mbuf_Prepend(). Adjust some spacing to make things more consistent.
1999-05-09 20:02:29 +00:00
/* 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;
1995-01-31 06:29:58 +00:00
#else
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
goto reqreject;
1995-01-31 06:29:58 +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);
1995-01-31 06:29:58 +00:00
break;
}
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
1995-01-31 06:29:58 +00:00
case TY_SDP:
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s\n", request);
This commit was manufactured by cvs2svn to create branch 'MP'.
1998-01-29 00:44:16 +00:00
switch (mode_type) {
case MODE_REQ:
case MODE_NAK:
case MODE_REJ:
break;
}
break;
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
case TY_CALLBACK:
if (length == 2)
op = CALLBACK_NONE;
else
op = (int)cp[2];
sz = length - 3;
switch (op) {
case CALLBACK_AUTH:
log_Printf(LogLCP, "%s Auth\n", request);
break;
case CALLBACK_DIALSTRING:
log_Printf(LogLCP, "%s Dialstring %.*s\n", request, sz, cp + 3);
break;
case CALLBACK_LOCATION:
log_Printf(LogLCP, "%s Location %.*s\n", request, sz, cp + 3);
break;
case CALLBACK_E164:
log_Printf(LogLCP, "%s E.164 (%.*s)\n", request, sz, cp + 3);
break;
case CALLBACK_NAME:
log_Printf(LogLCP, "%s Name %.*s\n", request, sz, cp + 3);
break;
case CALLBACK_CBCP:
log_Printf(LogLCP, "%s CBCP\n", request);
break;
default:
log_Printf(LogLCP, "%s ???\n", request);
break;
}
switch (mode_type) {
case MODE_REQ:
callback_req = 1;
if (p->type != PHYS_DIRECT)
goto reqreject;
if ((p->dl->cfg.callback.opmask & CALLBACK_BIT(op)) &&
(op != CALLBACK_AUTH || p->link.lcp.auth_ineed) &&
(op != CALLBACK_E164 ||
E164ok(&p->dl->cfg.callback, cp + 3, sz))) {
lcp->his_callback.opmask = CALLBACK_BIT(op);
if (sz > sizeof lcp->his_callback.msg - 1) {
sz = sizeof lcp->his_callback.msg - 1;
log_Printf(LogWARN, "Truncating option arg to %d octets\n", sz);
}
memcpy(lcp->his_callback.msg, cp + 3, sz);
lcp->his_callback.msg[sz] = '\0';
memcpy(dec->ackend, cp, sz + 3);
dec->ackend += sz + 3;
} else if ((p->dl->cfg.callback.opmask & CALLBACK_BIT(CALLBACK_AUTH)) &&
p->link.lcp.auth_ineed) {
*dec->nakend++ = *cp;
*dec->nakend++ = 3;
*dec->nakend++ = CALLBACK_AUTH;
} else if (p->dl->cfg.callback.opmask & CALLBACK_BIT(CALLBACK_CBCP)) {
*dec->nakend++ = *cp;
*dec->nakend++ = 3;
*dec->nakend++ = CALLBACK_CBCP;
} else if (p->dl->cfg.callback.opmask & CALLBACK_BIT(CALLBACK_E164)) {
*dec->nakend++ = *cp;
*dec->nakend++ = 3;
*dec->nakend++ = CALLBACK_E164;
} else if (p->dl->cfg.callback.opmask & CALLBACK_BIT(CALLBACK_AUTH)) {
log_Printf(LogWARN, "Cannot insist on auth callback without"
" PAP or CHAP enabled !\n");
*dec->nakend++ = *cp;
*dec->nakend++ = 2;
} else
goto reqreject;
break;
case MODE_NAK:
If we've configured CBCP *and* another not-so-important CALLBACK protocol and end up agreeing CBCP, DTRT and go into CBCP phase rather than mistakenly terminating as if CBCP wasn't agreed. Problem reported by: Alexander Dubinin <alex@nstl.nnov.ru>
1999-01-12 21:50:20 +00:00
/* We don't do what he NAKs with, we do things in our preferred order */
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
if (lcp->want_callback.opmask & CALLBACK_BIT(CALLBACK_AUTH))
lcp->want_callback.opmask &= ~CALLBACK_BIT(CALLBACK_AUTH);
else if (lcp->want_callback.opmask & CALLBACK_BIT(CALLBACK_CBCP))
lcp->want_callback.opmask &= ~CALLBACK_BIT(CALLBACK_CBCP);
else if (lcp->want_callback.opmask & CALLBACK_BIT(CALLBACK_E164))
lcp->want_callback.opmask &= ~CALLBACK_BIT(CALLBACK_E164);
if (lcp->want_callback.opmask == CALLBACK_BIT(CALLBACK_NONE)) {
log_Printf(LogPHASE, "Peer NAKd all callbacks, trying none\n");
lcp->want_callback.opmask = 0;
} else if (!lcp->want_callback.opmask) {
log_Printf(LogPHASE, "Peer NAKd last configured callback\n");
fsm_Close(&lcp->fsm);
}
break;
case MODE_REJ:
if (lcp->want_callback.opmask & CALLBACK_BIT(CALLBACK_NONE)) {
lcp->his_reject |= (1 << type);
lcp->want_callback.opmask = 0;
} else {
log_Printf(LogPHASE, "Peer rejected *required* callback\n");
fsm_Close(&lcp->fsm);
}
break;
}
break;
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
case TY_SHORTSEQ:
mp = &lcp->fsm.bundle->ncp.mp;
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s\n", request);
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
Don't pass global vars as args. Remove local/global conflicts.
1997-11-08 00:28:11 +00:00
switch (mode_type) {
1995-01-31 06:29:58 +00:00
case MODE_REQ:
o Don't go multilink if the peer hasn't asked for it o Don't go multilink if the peer hasn't agreed o When negotiating multilink, don't try to push the peers MRU up according to our `set mtu' setting. Instead, accept whatever MRU they want and push their MRRU up instead. o When *not* negotiating multilink, REJ short sequence number REQs.
1998-04-23 18:55:50 +00:00
if (lcp->want_mrru && IsAccepted(mp->cfg.shortseq)) {
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
lcp->his_shortseq = 1;
memcpy(dec->ackend, cp, length);
dec->ackend += length;
} else
goto reqreject;
break;
1995-01-31 06:29:58 +00:00
case MODE_NAK:
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
/*
* He's trying to get us to ask for short sequence numbers.
* We ignore the NAK and honour our configuration file instead.
*/
break;
1995-01-31 06:29:58 +00:00
case MODE_REJ:
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
lcp->his_reject |= (1 << type);
o Behave correctly when short sequence numbers are denied. o Determine if one seq is less than the other correctly, given the point at which they wrap (again).
1998-05-06 23:49:48 +00:00
lcp->want_shortseq = 0; /* For when we hit MP */
1995-01-31 06:29:58 +00:00
break;
}
break;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
case TY_ENDDISC:
Initialise `mp'.
1999-06-09 20:27:26 +00:00
mp = &lcp->fsm.bundle->ncp.mp;
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s %s\n", request,
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
mp_Enddisc(cp[2], cp + 3, length - 3));
switch (mode_type) {
case MODE_REQ:
o Defer setting up pap/chap based IP numbers & labels until after we've determined if we're going to join another ppp invocation. o Make ``show link'' show all link details, and ``show links'' just give a list of links and their current status. o Show our current label in ``show bundle''. o Allow link cloning and removal as soon as our MRRU is set. o Make ``show lcp'' require context as nothing will ever change in our MP LCP (it's auto-configured as per rfc1990). o Initialise our LQM owner in hdlc_Init(). o Store our endpoint discriminator and authentication name at both the datalink and multilink level and compare them when we've finished AUTHENTICATE phase and before entering NETWORK phase. If they don't match, close the link again. Display the information in the appropriate ``show'' command. o Initialise datalink::phone and datalink::fsmp.object properly when we're cloning the link. o Show which link we're passing LQRs on in our diagnostics. o Reject endpoint discriminator REQs at the logical multilink level. o Remove the rest of our CARRIER and LINK logging setup.
1998-04-24 19:16:15 +00:00
if (!p) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, " ENDDISC rejected - not a physical link\n");
o Defer setting up pap/chap based IP numbers & labels until after we've determined if we're going to join another ppp invocation. o Make ``show link'' show all link details, and ``show links'' just give a list of links and their current status. o Show our current label in ``show bundle''. o Allow link cloning and removal as soon as our MRRU is set. o Make ``show lcp'' require context as nothing will ever change in our MP LCP (it's auto-configured as per rfc1990). o Initialise our LQM owner in hdlc_Init(). o Store our endpoint discriminator and authentication name at both the datalink and multilink level and compare them when we've finished AUTHENTICATE phase and before entering NETWORK phase. If they don't match, close the link again. Display the information in the appropriate ``show'' command. o Initialise datalink::phone and datalink::fsmp.object properly when we're cloning the link. o Show which link we're passing LQRs on in our diagnostics. o Reject endpoint discriminator REQs at the logical multilink level. o Remove the rest of our CARRIER and LINK logging setup.
1998-04-24 19:16:15 +00:00
goto reqreject;
Allow our endpoint discriminator to be enabled, disabled, accepted and denied. This is necessary for some MP implementations that get confused if you accept their endpoint discriminator but reject their MRRU.
1999-06-09 16:54:04 +00:00
} else if (!IsAccepted(mp->cfg.negenddisc))
goto reqreject;
else if (length-3 < sizeof p->dl->peer.enddisc.address &&
o Defer setting up pap/chap based IP numbers & labels until after we've determined if we're going to join another ppp invocation. o Make ``show link'' show all link details, and ``show links'' just give a list of links and their current status. o Show our current label in ``show bundle''. o Allow link cloning and removal as soon as our MRRU is set. o Make ``show lcp'' require context as nothing will ever change in our MP LCP (it's auto-configured as per rfc1990). o Initialise our LQM owner in hdlc_Init(). o Store our endpoint discriminator and authentication name at both the datalink and multilink level and compare them when we've finished AUTHENTICATE phase and before entering NETWORK phase. If they don't match, close the link again. Display the information in the appropriate ``show'' command. o Initialise datalink::phone and datalink::fsmp.object properly when we're cloning the link. o Show which link we're passing LQRs on in our diagnostics. o Reject endpoint discriminator REQs at the logical multilink level. o Remove the rest of our CARRIER and LINK logging setup.
1998-04-24 19:16:15 +00:00
cp[2] <= MAX_ENDDISC_CLASS) {
p->dl->peer.enddisc.class = cp[2];
p->dl->peer.enddisc.len = length-3;
memcpy(p->dl->peer.enddisc.address, cp + 3, length - 3);
p->dl->peer.enddisc.address[length - 3] = '\0';
/* XXX: If mp->active, compare and NAK with mp->peer ? */
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
memcpy(dec->ackend, cp, length);
dec->ackend += length;
} else {
if (cp[2] > MAX_ENDDISC_CLASS)
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, " ENDDISC rejected - unrecognised class %d\n",
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
cp[2]);
else
o Fix remaining sizeof problems for 64 bit machines. o Allow ``set ....'' when we have multiple links but aren't in multilink mode. o Do a TLS when we receive a ``Open'' event in ``Closed'' state, despite the rfc state transition table. This is clearly an error in the RFC as TLS cannot have yet been called (without TLF) in the ``Closed'' state. I've posted a message to comp.protocols.ppp for confirmation.
1998-06-27 23:48:54 +00:00
log_Printf(LogLCP, " ENDDISC rejected - local max length is %ld\n",
(long)(sizeof p->dl->peer.enddisc.address - 1));
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
goto reqreject;
}
break;
Allow our endpoint discriminator to be enabled, disabled, accepted and denied. This is necessary for some MP implementations that get confused if you accept their endpoint discriminator but reject their MRRU.
1999-06-09 16:54:04 +00:00
case MODE_NAK: /* Treat this as a REJ, we don't vary our disc (yet) */
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
case MODE_REJ:
lcp->his_reject |= (1 << type);
break;
}
break;
1995-01-31 06:29:58 +00:00
default:
Cosmetic (style): sizeof(var) -> sizeof var sizeof type -> sizeof(type) Suggested by: J Wunsch <j@uriah.heep.sax.de>
1997-12-24 09:29:17 +00:00
sz = (sizeof desc - 2) / 2;
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
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]);
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "%s%s\n", request, desc);
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
Don't pass global vars as args. Remove local/global conflicts.
1997-11-08 00:28:11 +00:00
if (mode_type == MODE_REQ) {
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
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);
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogLCP, "Can't REJ length %d - trunating to %d\n",
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
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);
Abstract the CCP layer a level. Add DEFLATE support.
1997-12-03 10:23:54 +00:00
if (length != cp[1])
Negotiate MRRU, SHORTSEQ and ENDDISC. ENDDISC doesn't imply multilink ('cos I've seen my ISP REQ it without multilink). Setting MRRU is ifdef'd out until it's debugged and we can merge -direct links with other running programs. Fix MTU setting.
1998-04-23 03:23:03 +00:00
length = 0; /* force our way out of the loop */
New user Process PPP based on iij-ppp0.94beta2. o Supporting SYNC SIO device (But need a device driver) - add "set speed sync" o Fixing bug for Predictor-1 function. o Add new parameter that re-sent interval for set timeout commands. o Improving RTT (Round Trip Time) and reducing processor time. - Previous Timer service was using polling, and now using SIGALRM ;-) - A 0.94beta2 will not work correctly.... -- Follows are additinal feature not including 0.94beta2 o Support Proxy ARP - add "enable/disable proxy" commands o Marging common routine in CHAP/PAP. o Enhancing LCP/IPCP log information. o Support local Authfication connection on port 300x and tty. - You can set up pair of your "hostname -s" and password in ppp.secret. if either ppp.secret file nor your hostname line don't exist, It will notify a message and working as same as previous version.(Backword compatibility) - If you did set up them, It's allow connection but nothing to do except help and passwd command. - add "passwd yourpasswd" commands o Support afilter - keep Alive filter that a packet can send/receiving according to ifilter/ofilter but doesn't count it as preventing idle timer expires. - Same syntax of other filters. o Fixing bugs reported by current user for previous one. Thanks !! Reviewed by: Atsushi Murai (amurai@spec.co.jp)
1995-02-26 12:18:08 +00:00
}
1995-01-31 06:29:58 +00:00
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
Make gcc-2.8.1 build ppp cleanly. Support OpenBSD again.
1998-04-25 10:49:52 +00:00
if (mode_type != MODE_NOP) {
o Support callback types NONE, E.164, AUTH and CBCP. (see the new ``set callback'' and ``set cbcp'' commands) o Add a ``cbcp'' log level and mbuf type. o Don't dump core when \T is given in ``set login'' or ``set hangup''. o Allow ``*'' and blanks as placeholders in ppp.secret and allow a fifth field for specifying auth/cbcp dialback parameters. o Remove a few extraneous #includes o Define the default number of REQs (restart counter) in defs.h rather than hardcoding ``5'' all over the place. o Fix a few man page inconsistencies.
1998-08-07 18:42:51 +00:00
if (mode_type == MODE_REQ && p && p->type == PHYS_DIRECT &&
p->dl->cfg.callback.opmask && !callback_req &&
!(p->dl->cfg.callback.opmask & CALLBACK_BIT(CALLBACK_NONE))) {
/* We *REQUIRE* that the peer requests callback */
*dec->nakend++ = TY_CALLBACK;
*dec->nakend++ = 3;
if ((p->dl->cfg.callback.opmask & CALLBACK_BIT(CALLBACK_AUTH)) &&
p->link.lcp.auth_ineed)
*dec->nakend++ = CALLBACK_AUTH;
else if (p->dl->cfg.callback.opmask & CALLBACK_BIT(CALLBACK_CBCP))
*dec->nakend++ = CALLBACK_CBCP;
else if (p->dl->cfg.callback.opmask & CALLBACK_BIT(CALLBACK_E164))
*dec->nakend++ = CALLBACK_E164;
else {
log_Printf(LogWARN, "Cannot insist on auth callback without"
" PAP or CHAP enabled !\n");
dec->nakend[-1] = 2; /* XXX: Silly ! */
}
}
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 (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;
Make gcc-2.8.1 build ppp cleanly. Support OpenBSD again.
1998-04-25 10:49:52 +00:00
}
1995-01-31 06:29:58 +00:00
}
o Redesign the layering mechanism and make the aliasing code part of the layering. We now ``stack'' layers as soon as we open the device (when we figure out what we're dealing with). A static set of `dispatch' routines are also declared for dealing with incoming packets after they've been `pulled' up through the stacked layers. Physical devices are now assigned handlers based on the device type when they're opened. For the moment there are three device types; ttys, execs and tcps. o Increment version number to 2.2 o Make an entry in [uw]tmp for non-tty -direct invocations (after pap/chap authentication). o Make throughput counters quad_t's o Account for the absolute number of mbuf malloc()s and free()s in ``show mem''. o ``show modem'' becomes ``show physical''.
1999-05-08 11:07:56 +00:00
extern struct mbuf *
lcp_Input(struct bundle *bundle, struct link *l, struct mbuf *bp)
1995-01-31 06:29:58 +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 Alter the mbuf type as it's processed by different layers. o Show more information about missing MP fragments in ``show mp''. o Do away with mbuf_Log(). It was showing mbuf stats twice on receipt of LCP/CCP/IPCP packets.... ???!!? o Pre-allocate a bit extra when creating LQR packets to avoid having to allocate another mbuf in mbuf_Prepend().
1999-06-02 15:59:09 +00:00
mbuf_SetType(bp, MB_LCPIN);
o Redesign the layering mechanism and make the aliasing code part of the layering. We now ``stack'' layers as soon as we open the device (when we figure out what we're dealing with). A static set of `dispatch' routines are also declared for dealing with incoming packets after they've been `pulled' up through the stacked layers. Physical devices are now assigned handlers based on the device type when they're opened. For the moment there are three device types; ttys, execs and tcps. o Increment version number to 2.2 o Make an entry in [uw]tmp for non-tty -direct invocations (after pap/chap authentication). o Make throughput counters quad_t's o Account for the absolute number of mbuf malloc()s and free()s in ``show mem''. o ``show modem'' becomes ``show physical''.
1999-05-08 11:07:56 +00:00
fsm_Input(&l->lcp.fsm, bp);
return NULL;
1995-01-31 06:29:58 +00:00
}
Reference in New Issue Copy Permalink