d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
331c488d69
freebsd-dev/usr.sbin/ppp/mp.c

1196 lines
33 KiB
C
Raw Normal View History

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
/*-
* Copyright (c) 1998 Brian Somers <brian@Awfulhak.org>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
$Id$ -> $FreeBSD$
1999-08-28 01:35:59 +00:00
* $FreeBSD$
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
*/
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>
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
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.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 <arpa/inet.h>
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
#include <net/if_dl.h>
#include <sys/socket.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>
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
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
#include <errno.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 <paths.h>
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
#include <stdlib.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 <stdio.h>
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
#include <string.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/stat.h>
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
#include <termios.h>
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
#include <unistd.h>
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
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"
o Add the -foreground switch. This switch behaves like -background except that ppp stays in the foreground. o Add the -quiet switch to quieten ppps startup o Add the -nat flag and discourage the use of the -alias flag. Both do the same thing. o Correct some nat usage strings. o Change the internal ``alias'' command to ``nat''.
1999-08-19 18:15:52 +00:00
#ifndef NONAT
Cosmetic: alias_cmd -> nat_cmd after a repo-copy
1999-09-08 07:34:52 +00:00
#include "nat_cmd.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
#endif
#include "vjcomp.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"
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
#include "command.h"
#include "mbuf.h"
#include "log.h"
#include "timer.h"
#include "fsm.h"
#include "iplist.h"
#include "throughput.h"
#include "slcompress.h"
Put the IP buffer queues into struct ipcp. Forgotten by: me
1998-08-26 17:39:37 +00:00
#include "lqr.h"
#include "hdlc.h"
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
#include "ipcp.h"
#include "auth.h"
#include "lcp.h"
#include "async.h"
#include "ccp.h"
#include "link.h"
#include "descriptor.h"
#include "physical.h"
#include "chat.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"
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
#include "filter.h"
#include "mp.h"
#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 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
#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
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
#include "bundle.h"
#include "ip.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 "prompt.h"
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
#include "id.h"
#include "arp.h"
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
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
void
peerid_Init(struct peerid *peer)
{
peer->enddisc.class = 0;
*peer->enddisc.address = '\0';
peer->enddisc.len = 0;
*peer->authname = '\0';
}
int
peerid_Equal(const struct peerid *p1, const struct peerid *p2)
{
return !strcmp(p1->authname, p2->authname) &&
p1->enddisc.class == p2->enddisc.class &&
p1->enddisc.len == p2->enddisc.len &&
!memcmp(p1->enddisc.address, p2->enddisc.address, p1->enddisc.len);
}
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
static u_int32_t
o Increment expected MP fragment numbers correctly, according to SHORTSEQ negotiation. o Don't forget to attach incoming fragments with a number greater than everything else in the queue (rather than leaking memory). o Output the link name with the ``other'' hdlc diagnostic message. o Correct a VJ diagnostic (`COMPPROTO', not `proto').
1998-05-04 03:00:09 +00:00
inc_seq(unsigned is12bit, u_int32_t seq)
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
{
seq++;
o Increment expected MP fragment numbers correctly, according to SHORTSEQ negotiation. o Don't forget to attach incoming fragments with a number greater than everything else in the queue (rather than leaking memory). o Output the link name with the ``other'' hdlc diagnostic message. o Correct a VJ diagnostic (`COMPPROTO', not `proto').
1998-05-04 03:00:09 +00:00
if (is12bit) {
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
if (seq & 0xfffff000)
seq = 0;
} else if (seq & 0xff000000)
seq = 0;
return seq;
}
Allow a 1000 fragment sliding window when placing a new fragment in our inbound multilnk queue. This fixes wrapping problems when something arrives out-of-sequence.
1998-05-04 21:42:41 +00:00
static int
isbefore(unsigned is12bit, u_int32_t seq1, u_int32_t seq2)
{
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
u_int32_t max = (is12bit ? 0xfff : 0xffffff) - 0x200;
if (seq1 > max) {
if (seq2 < 0x200 || seq2 > seq1)
return 1;
} else if ((seq1 > 0x200 || seq2 <= max) && seq1 < seq2)
return 1;
return 0;
Allow a 1000 fragment sliding window when placing a new fragment in our inbound multilnk queue. This fixes wrapping problems when something arrives out-of-sequence.
1998-05-04 21:42:41 +00:00
}
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
static int
mp_ReadHeader(struct mp *mp, struct mbuf *m, struct mp_header *header)
{
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 (mp->local_is12bit) {
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 val;
ua_ntohs(MBUF_CTOP(m), &val);
if (val & 0x3000) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogWARN, "Oops - MP header without required zero bits\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
return 0;
}
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
header->begin = val & 0x8000 ? 1 : 0;
header->end = val & 0x4000 ? 1 : 0;
header->seq = val & 0x0fff;
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
return 2;
} else {
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(MBUF_CTOP(m), &header->seq);
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
if (header->seq & 0x3f000000) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogWARN, "Oops - MP header without required zero bits\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
return 0;
}
o Add ipcp_SetLink() for attaching IPCP to the correct link, and call it after link authentication. o Pretty print our bundle MTU. o Correct MP header encoding and decoding (should be network byte order). o Add some debug diagnostics so that we can see MP fragment sending and re-assembly.
1998-04-23 18:56:21 +00:00
header->begin = header->seq & 0x80000000 ? 1 : 0;
header->end = header->seq & 0x40000000 ? 1 : 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
header->seq &= 0x00ffffff;
return 4;
}
}
static void
mp_LayerStart(void *v, struct fsm *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
/* The given FSM (ccp) is about to start up ! */
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
}
static void
mp_LayerUp(void *v, struct fsm *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
/* The given fsm (ccp) is now up */
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
}
static void
mp_LayerDown(void *v, struct fsm *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
/* The given FSM (ccp) has been told to come down */
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
}
static void
mp_LayerFinish(void *v, struct fsm *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
/* The given fsm (ccp) is now down */
o Activate link-level CCPs in multilink mode, by bringing them into the ST_STOPPED state. o Allow an optional ccp|lcp argument to `down'. The default is still lcp (as before). You can now call down with no context in multilink mode, in which case it'll down the multilink ccp or the entire bundle (*very* rude). o Allow an optional `!' after `close ccp' (close ccp!) to tell ccp to stay in the CLOSED state after the terminate ACK. The default is now to re-enter STOPPED so that the peer can bring the layer back up if desired. o Always handle proto-compressed packets, even if we've agreed (in LCP) that the peer will not send us 1 byte protocols. If the peer violates the LCP agreement, log it to the HDLC log. o Fix some comments.
1998-05-15 18:21:12 +00:00
if (fp->state == ST_CLOSED && fp->open_mode == OPEN_PASSIVE)
fsm_Open(fp); /* CCP goes to ST_STOPPED */
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
}
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
static void
mp_UpDown(void *v)
{
struct mp *mp = (struct mp *)v;
int percent;
Maintain input and output throughput averages and choose the highest of the two when calculating the MP throughput average for the ``set autoload'' implementation. This makes more sense as all links I know of are full-duplex. This also means that people may need to adjust their autoload settings as 100% bandwidth is now the theoretical maximum rather than 200% (but of course, halfing the current settings is probably not the correct answer either!). This involves a ppp version bump as we need to pass an extra throughput array through the MP local domain socket.
2000-08-15 10:25:42 +00:00
percent = MAX(mp->link.stats.total.in.OctetsPerSecond,
mp->link.stats.total.out.OctetsPerSecond) * 800 /
mp->bundle->bandwidth;
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
if (percent >= mp->cfg.autoload.max) {
log_Printf(LogDEBUG, "%d%% saturation - bring a link up ?\n", percent);
bundle_AutoAdjust(mp->bundle, percent, AUTO_UP);
} else if (percent <= mp->cfg.autoload.min) {
log_Printf(LogDEBUG, "%d%% saturation - bring a link down ?\n", percent);
bundle_AutoAdjust(mp->bundle, percent, AUTO_DOWN);
}
}
void
mp_StopAutoloadTimer(struct mp *mp)
{
Calculate the average link throughput using a counter based on the cumulative total of all active links rather than basing it on the total of PROTO_MP traffic. This fixes a problem whereby Cisco routers send PROTO_IP packets only when there's only one link (hmm, what a good idea!).
2000-08-15 00:59:21 +00:00
throughput_stop(&mp->link.stats.total);
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
}
void
mp_CheckAutoloadTimer(struct mp *mp)
{
Calculate the average link throughput using a counter based on the cumulative total of all active links rather than basing it on the total of PROTO_MP traffic. This fixes a problem whereby Cisco routers send PROTO_IP packets only when there's only one link (hmm, what a good idea!).
2000-08-15 00:59:21 +00:00
if (mp->link.stats.total.SamplePeriod != mp->cfg.autoload.period) {
throughput_destroy(&mp->link.stats.total);
throughput_init(&mp->link.stats.total, mp->cfg.autoload.period);
throughput_callback(&mp->link.stats.total, mp_UpDown, mp);
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
}
if (bundle_WantAutoloadTimer(mp->bundle))
Calculate the average link throughput using a counter based on the cumulative total of all active links rather than basing it on the total of PROTO_MP traffic. This fixes a problem whereby Cisco routers send PROTO_IP packets only when there's only one link (hmm, what a good idea!).
2000-08-15 00:59:21 +00:00
throughput_start(&mp->link.stats.total, "MP throughput", 1);
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
else
mp_StopAutoloadTimer(mp);
}
void
mp_RestartAutoloadTimer(struct mp *mp)
{
Calculate the average link throughput using a counter based on the cumulative total of all active links rather than basing it on the total of PROTO_MP traffic. This fixes a problem whereby Cisco routers send PROTO_IP packets only when there's only one link (hmm, what a good idea!).
2000-08-15 00:59:21 +00:00
if (mp->link.stats.total.SamplePeriod != mp->cfg.autoload.period)
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
mp_CheckAutoloadTimer(mp);
else
Calculate the average link throughput using a counter based on the cumulative total of all active links rather than basing it on the total of PROTO_MP traffic. This fixes a problem whereby Cisco routers send PROTO_IP packets only when there's only one link (hmm, what a good idea!).
2000-08-15 00:59:21 +00:00
throughput_clear(&mp->link.stats.total, THROUGHPUT_OVERALL, NULL);
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
}
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
void
mp_Init(struct mp *mp, struct bundle *bundle)
{
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->peer_is12bit = mp->local_is12bit = 0;
mp->peer_mrru = mp->local_mrru = 0;
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
peerid_Init(&mp->peer);
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
mp->out.seq = 0;
mp->out.link = 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
mp->seq.min_in = 0;
mp->seq.next_in = 0;
mp->inbufs = NULL;
mp->bundle = bundle;
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
mp->link.type = LOGICAL_LINK;
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
mp->link.name = "mp";
mp->link.len = sizeof *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
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
mp->cfg.autoload.period = SAMPLE_PERIOD;
mp->cfg.autoload.min = mp->cfg.autoload.max = 0;
Calculate the average link throughput using a counter based on the cumulative total of all active links rather than basing it on the total of PROTO_MP traffic. This fixes a problem whereby Cisco routers send PROTO_IP packets only when there's only one link (hmm, what a good idea!).
2000-08-15 00:59:21 +00:00
throughput_init(&mp->link.stats.total, mp->cfg.autoload.period);
throughput_callback(&mp->link.stats.total, mp_UpDown, mp);
mp->link.stats.parent = NULL;
mp->link.stats.gather = 0; /* Let the physical links gather stats */
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
memset(mp->link.Queue, '\0', sizeof mp->link.Queue);
memset(mp->link.proto_in, '\0', sizeof mp->link.proto_in);
memset(mp->link.proto_out, '\0', sizeof mp->link.proto_out);
mp->fsmp.LayerStart = mp_LayerStart;
mp->fsmp.LayerUp = mp_LayerUp;
mp->fsmp.LayerDown = mp_LayerDown;
mp->fsmp.LayerFinish = mp_LayerFinish;
mp->fsmp.object = mp;
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
mpserver_Init(&mp->server);
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->cfg.mrru = 0;
mp->cfg.shortseq = NEG_ENABLED|NEG_ACCEPTED;
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
mp->cfg.negenddisc = NEG_ENABLED|NEG_ACCEPTED;
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->cfg.enddisc.class = 0;
*mp->cfg.enddisc.address = '\0';
mp->cfg.enddisc.len = 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_Init(&mp->link.lcp, mp->bundle, &mp->link, NULL);
ccp_Init(&mp->link.ccp, mp->bundle, &mp->link, &mp->fsmp);
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
link_EmptyStack(&mp->link);
link_Stack(&mp->link, &protolayer);
link_Stack(&mp->link, &ccplayer);
link_Stack(&mp->link, &vjlayer);
o Add the -foreground switch. This switch behaves like -background except that ppp stays in the foreground. o Add the -quiet switch to quieten ppps startup o Add the -nat flag and discourage the use of the -alias flag. Both do the same thing. o Correct some nat usage strings. o Change the internal ``alias'' command to ``nat''.
1999-08-19 18:15:52 +00:00
#ifndef NONAT
link_Stack(&mp->link, &natlayer);
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
#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
}
int
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
mp_Up(struct mp *mp, struct datalink *dl)
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 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
struct lcp *lcp = &dl->physical->link.lcp;
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 (mp->active) {
/* We're adding a link - do a last validation on our parameters */
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
if (!peerid_Equal(&dl->peer, &mp->peer)) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogPHASE, "%s: Inappropriate peer !\n", dl->name);
Be more verbose when a second link doesn't match the first because the peer enddisc/authname is different.
2000-05-22 08:23:13 +00:00
log_Printf(LogPHASE, " Attached to peer %s/%s\n", mp->peer.authname,
mp_Enddisc(mp->peer.enddisc.class, mp->peer.enddisc.address,
mp->peer.enddisc.len));
log_Printf(LogPHASE, " New link is peer %s/%s\n", dl->peer.authname,
mp_Enddisc(dl->peer.enddisc.class, dl->peer.enddisc.address,
dl->peer.enddisc.len));
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
return MP_FAILED;
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
}
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
if (mp->local_mrru != lcp->want_mrru ||
mp->peer_mrru != lcp->his_mrru ||
mp->local_is12bit != lcp->want_shortseq ||
mp->peer_is12bit != lcp->his_shortseq) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogPHASE, "%s: Invalid MRRU/SHORTSEQ MP parameters !\n",
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
dl->name);
return MP_FAILED;
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 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
return MP_ADDED;
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
} else {
/* First link in multilink mode */
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
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
mp->local_mrru = lcp->want_mrru;
mp->peer_mrru = lcp->his_mrru;
mp->local_is12bit = lcp->want_shortseq;
mp->peer_is12bit = lcp->his_shortseq;
mp->peer = dl->peer;
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
Calculate the average link throughput using a counter based on the cumulative total of all active links rather than basing it on the total of PROTO_MP traffic. This fixes a problem whereby Cisco routers send PROTO_IP packets only when there's only one link (hmm, what a good idea!).
2000-08-15 00:59:21 +00:00
throughput_destroy(&mp->link.stats.total);
throughput_init(&mp->link.stats.total, mp->cfg.autoload.period);
throughput_callback(&mp->link.stats.total, mp_UpDown, mp);
o Initialise MP correctly depending on if we're the first link in mp_Up(). o Bring MP and its CCP down when we enter phase TERMINATE, and ditch everything in the incoming packet queue. o Enable MRRU negotiation. Now, we can multilink mode, but only with one physical link. o Close the link if the peer PROTO REJs PROTO_MP. o Prepend our protocol before passing a packet to struct mp for fragmentation. o Log info messages to DEBUG, not ERROR (oops). o Align `show mp' output (again).
1998-04-23 21:50:13 +00:00
memset(mp->link.Queue, '\0', sizeof mp->link.Queue);
memset(mp->link.proto_in, '\0', sizeof mp->link.proto_in);
memset(mp->link.proto_out, '\0', sizeof mp->link.proto_out);
Calculate the average link throughput using a counter based on the cumulative total of all active links rather than basing it on the total of PROTO_MP traffic. This fixes a problem whereby Cisco routers send PROTO_IP packets only when there's only one link (hmm, what a good idea!).
2000-08-15 00:59:21 +00:00
/* Tell the link who it belongs to */
dl->physical->link.stats.parent = &mp->link.stats.total;
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
mp->out.seq = 0;
mp->out.link = 0;
o Initialise MP correctly depending on if we're the first link in mp_Up(). o Bring MP and its CCP down when we enter phase TERMINATE, and ditch everything in the incoming packet queue. o Enable MRRU negotiation. Now, we can multilink mode, but only with one physical link. o Close the link if the peer PROTO REJs PROTO_MP. o Prepend our protocol before passing a packet to struct mp for fragmentation. o Log info messages to DEBUG, not ERROR (oops). o Align `show mp' output (again).
1998-04-23 21:50:13 +00:00
mp->seq.min_in = 0;
mp->seq.next_in = 0;
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
/*
* Now we create our server socket.
* If it already exists, join it. Otherwise, create and own it
*/
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
switch (mpserver_Open(&mp->server, &mp->peer)) {
case MPSERVER_CONNECTED:
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogPHASE, "mp: Transfer link on %s\n",
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
mp->server.socket.sun_path);
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
mp->server.send.dl = dl; /* Defer 'till it's safe to send */
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
return MP_LINKSENT;
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
case MPSERVER_FAILED:
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
return MP_FAILED;
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
case MPSERVER_LISTENING:
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogPHASE, "mp: Listening on %s\n", mp->server.socket.sun_path);
log_Printf(LogPHASE, " First link: %s\n", dl->name);
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
/* Re-point our IPCP layer at our MP link */
ipcp_SetLink(&mp->bundle->ncp.ipcp, &mp->link);
/* Our lcp's already up 'cos of the NULL parent */
The CCP layer now behaves as follows: o If we've denied and disabled all compression protocols, stay in ST_INITIAL and do an LCP protocol reject if we receive any CCP packets. o If we've disabled all compression protocols, go to ST_STOPPED and wait for the other side to ask for something. o If we've got anything enabled, start REQing as soon as the auth layer is up. o If we're in multilink mode, than the link level CCP goes straight to ST_STOPPED irrespective of what's configured so that we never try to compress compressed stuff by default.
1998-06-30 23:04:17 +00:00
if (ccp_SetOpenMode(&mp->link.ccp)) {
fsm_Up(&mp->link.ccp.fsm);
fsm_Open(&mp->link.ccp.fsm);
}
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
mp->active = 1;
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
break;
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
}
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 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
return MP_UP;
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
}
o Initialise MP correctly depending on if we're the first link in mp_Up(). o Bring MP and its CCP down when we enter phase TERMINATE, and ditch everything in the incoming packet queue. o Enable MRRU negotiation. Now, we can multilink mode, but only with one physical link. o Close the link if the peer PROTO REJs PROTO_MP. o Prepend our protocol before passing a packet to struct mp for fragmentation. o Log info messages to DEBUG, not ERROR (oops). o Align `show mp' output (again).
1998-04-23 21:50:13 +00:00
void
mp_Down(struct mp *mp)
{
if (mp->active) {
struct mbuf *next;
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
/* Stop that ! */
mp_StopAutoloadTimer(mp);
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
/* Don't want any more of these */
mpserver_Close(&mp->server);
/* CCP goes down with a bang */
Create & use fsm2initial(), a function to bring a state machine back to ST_INITIAL without going through any unnecessary TLS/TLF pairs.
1998-06-20 00:19:42 +00:00
fsm2initial(&mp->link.ccp.fsm);
o Initialise MP correctly depending on if we're the first link in mp_Up(). o Bring MP and its CCP down when we enter phase TERMINATE, and ditch everything in the incoming packet queue. o Enable MRRU negotiation. Now, we can multilink mode, but only with one physical link. o Close the link if the peer PROTO REJs PROTO_MP. o Prepend our protocol before passing a packet to struct mp for fragmentation. o Log info messages to DEBUG, not ERROR (oops). o Align `show mp' output (again).
1998-04-23 21:50:13 +00:00
/* Received fragments go in the bit-bucket */
while (mp->inbufs) {
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
next = mp->inbufs->m_nextpkt;
m_freem(mp->inbufs);
o Initialise MP correctly depending on if we're the first link in mp_Up(). o Bring MP and its CCP down when we enter phase TERMINATE, and ditch everything in the incoming packet queue. o Enable MRRU negotiation. Now, we can multilink mode, but only with one physical link. o Close the link if the peer PROTO REJs PROTO_MP. o Prepend our protocol before passing a packet to struct mp for fragmentation. o Log info messages to DEBUG, not ERROR (oops). o Align `show mp' output (again).
1998-04-23 21:50:13 +00:00
mp->inbufs = next;
}
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
peerid_Init(&mp->peer);
o Initialise MP correctly depending on if we're the first link in mp_Up(). o Bring MP and its CCP down when we enter phase TERMINATE, and ditch everything in the incoming packet queue. o Enable MRRU negotiation. Now, we can multilink mode, but only with one physical link. o Close the link if the peer PROTO REJs PROTO_MP. o Prepend our protocol before passing a packet to struct mp for fragmentation. o Log info messages to DEBUG, not ERROR (oops). o Align `show mp' output (again).
1998-04-23 21:50:13 +00:00
mp->active = 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
void
mp_linkInit(struct mp_link *mplink)
{
mplink->seq = 0;
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
mplink->bandwidth = 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
}
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
static void
mp_Assemble(struct mp *mp, struct mbuf *m, struct physical *p)
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
{
struct mp_header mh, h;
struct mbuf *q, *last;
int32_t seq;
o Make sure we adjust our min seq and process any outstanding queued incoming fragments when a link goes down. o Don't use the minimum sequence numbers of links that aren't open. o Understand sequence number wrapping when determining the minimum sequence number. o Add & adjust a few comments.
1998-05-23 17:05:28 +00:00
/*
* When `m' and `p' are NULL, it means our oldest link has gone down.
* We want to determine a new min, and process any intermediate stuff
* as normal
*/
if (m && mp_ReadHeader(mp, m, &mh) == 0) {
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
m_freem(m);
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
return;
}
o Make sure we adjust our min seq and process any outstanding queued incoming fragments when a link goes down. o Don't use the minimum sequence numbers of links that aren't open. o Understand sequence number wrapping when determining the minimum sequence number. o Add & adjust a few comments.
1998-05-23 17:05:28 +00:00
if (p) {
seq = p->dl->mp.seq;
p->dl->mp.seq = mh.seq;
} else
seq = mp->seq.min_in;
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
if (mp->seq.min_in == seq) {
/*
* We've received new data on the link that has our min (oldest) seq.
* Figure out which link now has the smallest (oldest) seq.
*/
struct datalink *dl;
o Make sure we adjust our min seq and process any outstanding queued incoming fragments when a link goes down. o Don't use the minimum sequence numbers of links that aren't open. o Understand sequence number wrapping when determining the minimum sequence number. o Add & adjust a few comments.
1998-05-23 17:05:28 +00:00
mp->seq.min_in = (u_int32_t)-1;
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
for (dl = mp->bundle->links; dl; dl = dl->next)
o Make sure we adjust our min seq and process any outstanding queued incoming fragments when a link goes down. o Don't use the minimum sequence numbers of links that aren't open. o Understand sequence number wrapping when determining the minimum sequence number. o Add & adjust a few comments.
1998-05-23 17:05:28 +00:00
if (dl->state == DATALINK_OPEN &&
(mp->seq.min_in == -1 ||
isbefore(mp->local_is12bit, dl->mp.seq, mp->seq.min_in)))
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
mp->seq.min_in = dl->mp.seq;
}
/*
* Now process as many of our fragments as we can, adding our new
* fragment in as we go, and ordering with the oldest at the top of
* the queue.
*/
last = NULL;
seq = mp->seq.next_in;
q = mp->inbufs;
Fix a rather nasty latency problem that occurs with single tcp sessions thorough an MP setup with only a single link.
2000-07-12 15:08:03 +00:00
while (q || m) {
if (!q) {
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
if (last)
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
last->m_nextpkt = m;
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
else
mp->inbufs = m;
q = m;
m = NULL;
Fix a rather nasty latency problem that occurs with single tcp sessions thorough an MP setup with only a single link.
2000-07-12 15:08:03 +00:00
h = mh;
} else {
mp_ReadHeader(mp, q, &h);
if (m && isbefore(mp->local_is12bit, mh.seq, h.seq)) {
/* Our received fragment fits in before this one, so link it in */
if (last)
last->m_nextpkt = m;
else
mp->inbufs = m;
m->m_nextpkt = q;
q = m;
h = mh;
m = NULL;
}
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
}
if (h.seq != seq) {
/* we're missing something :-( */
Fix some MP sequence number comparison bogons that are tickled by having different speed links in a bundle. This would manifest itself by having the link occasionally hang, but revive when a new connection is made.... Make ``show mp'' a bit prettier.
1999-06-03 13:29:32 +00:00
if (isbefore(mp->local_is12bit, seq, mp->seq.min_in)) {
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
/* we're never gonna get it */
struct mbuf *next;
/* Zap all older fragments */
while (mp->inbufs != q) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogDEBUG, "Drop frag\n");
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
next = mp->inbufs->m_nextpkt;
m_freem(mp->inbufs);
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
mp->inbufs = next;
}
/*
* Zap everything until the next `end' fragment OR just before
* the next `begin' fragment OR 'till seq.min_in - whichever
* comes first.
*/
do {
mp_ReadHeader(mp, mp->inbufs, &h);
if (h.begin) {
o Add ipcp_SetLink() for attaching IPCP to the correct link, and call it after link authentication. o Pretty print our bundle MTU. o Correct MP header encoding and decoding (should be network byte order). o Add some debug diagnostics so that we can see MP fragment sending and re-assembly.
1998-04-23 18:56:21 +00:00
/* We might be able to process this ! */
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
h.seq--; /* We're gonna look for fragment with h.seq+1 */
break;
}
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
next = mp->inbufs->m_nextpkt;
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogDEBUG, "Drop frag %u\n", h.seq);
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
m_freem(mp->inbufs);
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
mp->inbufs = next;
Fix some MP sequence number comparison bogons that are tickled by having different speed links in a bundle. This would manifest itself by having the link occasionally hang, but revive when a new connection is made.... Make ``show mp'' a bit prettier.
1999-06-03 13:29:32 +00:00
} while (mp->inbufs && (isbefore(mp->local_is12bit, mp->seq.min_in,
h.seq) || h.end));
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
/*
* Continue processing things from here.
* This deals with the possibility that we received a fragment
* on the slowest link that invalidates some of our data (because
* of the hole at `q'), but where there are subsequent `whole'
* packets that have already been received.
*/
o Increment expected MP fragment numbers correctly, according to SHORTSEQ negotiation. o Don't forget to attach incoming fragments with a number greater than everything else in the queue (rather than leaking memory). o Output the link name with the ``other'' hdlc diagnostic message. o Correct a VJ diagnostic (`COMPPROTO', not `proto').
1998-05-04 03:00:09 +00:00
mp->seq.next_in = seq = inc_seq(mp->local_is12bit, h.seq);
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
last = NULL;
q = mp->inbufs;
} else
/* we may still receive the missing fragment */
break;
} else if (h.end) {
/* We've got something, reassemble */
struct mbuf **frag = &q;
int len;
o Add ipcp_SetLink() for attaching IPCP to the correct link, and call it after link authentication. o Pretty print our bundle MTU. o Correct MP header encoding and decoding (should be network byte order). o Add some debug diagnostics so that we can see MP fragment sending and re-assembly.
1998-04-23 18:56:21 +00:00
u_long first = -1;
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
do {
*frag = mp->inbufs;
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
mp->inbufs = mp->inbufs->m_nextpkt;
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
len = mp_ReadHeader(mp, *frag, &h);
o Add ipcp_SetLink() for attaching IPCP to the correct link, and call it after link authentication. o Pretty print our bundle MTU. o Correct MP header encoding and decoding (should be network byte order). o Add some debug diagnostics so that we can see MP fragment sending and re-assembly.
1998-04-23 18:56:21 +00:00
if (first == -1)
first = h.seq;
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
(*frag)->m_offset += len;
(*frag)->m_len -= len;
(*frag)->m_nextpkt = NULL;
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
if (frag == &q && !h.begin) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogWARN, "Oops - MP frag %lu should have a begin flag\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
(u_long)h.seq);
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
m_freem(q);
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
q = NULL;
} else if (frag != &q && h.begin) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogWARN, "Oops - MP frag %lu should have an end flag\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
(u_long)h.seq - 1);
/*
* Stuff our fragment back at the front of the queue and zap
* our half-assembed packet.
*/
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
(*frag)->m_nextpkt = mp->inbufs;
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
mp->inbufs = *frag;
*frag = NULL;
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
m_freem(q);
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
q = NULL;
frag = &q;
h.end = 0; /* just in case it's a whole packet */
} else
do
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
frag = &(*frag)->m_next;
o Add ipcp_SetLink() for attaching IPCP to the correct link, and call it after link authentication. o Pretty print our bundle MTU. o Correct MP header encoding and decoding (should be network byte order). o Add some debug diagnostics so that we can see MP fragment sending and re-assembly.
1998-04-23 18:56:21 +00:00
while (*frag != NULL);
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
} while (!h.end);
if (q) {
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
q = m_pullup(q);
Allow ``host:port/udp'' devices and support ``host:port/tcp'' as being the same as the previous (still supported) ``host:port'' syntax for tcp socket devices. A udp device uses synchronous ppp rather than async, and avoids the double-retransmit overhead that comes with ppp over tcp (it's usually a bad idea to transport IP over a reliable transport that itself is using an unreliable transport). PPP over UDP provides througput of ** 1.5Mb per second ** with all compression disabled, maxing out a PPro/200 when running ppp twice, back-to-back. This proves that PPPoE is plausable in userland.... This change adds a few more handler functions to struct device and allows derivations of struct device (which may contain their own data etc) to pass themselves through the unix domain socket for MP. ** At last **, struct physical has lost all the tty crud ! iov2physical() is now smart enough to restore the correct stack of layers so that MP servers will work again. The version number has bumped as our MP link transfer contents have changed (they now may contain a `struct device'). Don't extract the protocol twice in MP mode (resulting in protocol rejects for every MP packet). This was broken with my original layering changes. Add ``Physical'' and ``Sync'' log levels for logging the relevent raw packets and add protocol-tracking LogDEBUG stuff in various LayerPush & LayerPull functions. Assign our physical device name for incoming tcp connections by calling getpeername(). Assign our physical device name for incoming udp connections from the address retrieved by the first recvfrom().
1999-05-12 09:49:12 +00:00
log_Printf(LogDEBUG, "MP: Reassembled frags %ld-%lu, length %d\n",
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
first, (u_long)h.seq, m_length(q));
link_PullPacket(&mp->link, MBUF_CTOP(q), q->m_len, mp->bundle);
m_freem(q);
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
}
o Increment expected MP fragment numbers correctly, according to SHORTSEQ negotiation. o Don't forget to attach incoming fragments with a number greater than everything else in the queue (rather than leaking memory). o Output the link name with the ``other'' hdlc diagnostic message. o Correct a VJ diagnostic (`COMPPROTO', not `proto').
1998-05-04 03:00:09 +00:00
mp->seq.next_in = seq = inc_seq(mp->local_is12bit, h.seq);
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
last = NULL;
q = mp->inbufs;
} else {
/* Look for the next fragment */
o Increment expected MP fragment numbers correctly, according to SHORTSEQ negotiation. o Don't forget to attach incoming fragments with a number greater than everything else in the queue (rather than leaking memory). o Output the link name with the ``other'' hdlc diagnostic message. o Correct a VJ diagnostic (`COMPPROTO', not `proto').
1998-05-04 03:00:09 +00:00
seq = inc_seq(mp->local_is12bit, seq);
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
last = q;
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
q = q->m_nextpkt;
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
}
}
if (m) {
/* We still have to find a home for our new fragment */
last = NULL;
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
for (q = mp->inbufs; q; last = q, q = q->m_nextpkt) {
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
mp_ReadHeader(mp, q, &h);
Allow a 1000 fragment sliding window when placing a new fragment in our inbound multilnk queue. This fixes wrapping problems when something arrives out-of-sequence.
1998-05-04 21:42:41 +00:00
if (isbefore(mp->local_is12bit, mh.seq, h.seq))
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
break;
}
o Increment expected MP fragment numbers correctly, according to SHORTSEQ negotiation. o Don't forget to attach incoming fragments with a number greater than everything else in the queue (rather than leaking memory). o Output the link name with the ``other'' hdlc diagnostic message. o Correct a VJ diagnostic (`COMPPROTO', not `proto').
1998-05-04 03:00:09 +00:00
/* Our received fragment fits in here */
if (last)
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
last->m_nextpkt = m;
o Increment expected MP fragment numbers correctly, according to SHORTSEQ negotiation. o Don't forget to attach incoming fragments with a number greater than everything else in the queue (rather than leaking memory). o Output the link name with the ``other'' hdlc diagnostic message. o Correct a VJ diagnostic (`COMPPROTO', not `proto').
1998-05-04 03:00:09 +00:00
else
mp->inbufs = m;
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
m->m_nextpkt = q;
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
}
}
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
struct mbuf *
mp_Input(struct bundle *bundle, struct link *l, struct mbuf *bp)
{
struct physical *p = link2physical(l);
if (!bundle->ncp.mp.active)
/* Let someone else deal with it ! */
return bp;
if (p == NULL) {
log_Printf(LogWARN, "DecodePacket: Can't do MP inside MP !\n");
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
m_freem(bp);
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
} else {
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
m_settype(bp, MB_MPIN);
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
mp_Assemble(&bundle->ncp.mp, bp, 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
}
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
return NULL;
}
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
static void
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
mp_Output(struct mp *mp, struct bundle *bundle, struct link *l,
struct mbuf *m, u_int32_t begin, u_int32_t end)
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
{
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
char prepend[4];
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
o Add ipcp_SetLink() for attaching IPCP to the correct link, and call it after link authentication. o Pretty print our bundle MTU. o Correct MP header encoding and decoding (should be network byte order). o Add some debug diagnostics so that we can see MP fragment sending and re-assembly.
1998-04-23 18:56:21 +00:00
/* Stuff an MP header on the front of our packet and send it */
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +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 (mp->peer_is12bit) {
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 val;
o Add ipcp_SetLink() for attaching IPCP to the correct link, and call it after link authentication. o Pretty print our bundle MTU. o Correct MP header encoding and decoding (should be network byte order). o Add some debug diagnostics so that we can see MP fragment sending and re-assembly.
1998-04-23 18:56:21 +00:00
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
val = (begin << 15) | (end << 14) | (u_int16_t)mp->out.seq;
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
ua_htons(&val, prepend);
m = m_prepend(m, prepend, 2, 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
} else {
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_int32_t val;
o Add ipcp_SetLink() for attaching IPCP to the correct link, and call it after link authentication. o Pretty print our bundle MTU. o Correct MP header encoding and decoding (should be network byte order). o Add some debug diagnostics so that we can see MP fragment sending and re-assembly.
1998-04-23 18:56:21 +00:00
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
val = (begin << 31) | (end << 30) | (u_int32_t)mp->out.seq;
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
ua_htonl(&val, prepend);
m = m_prepend(m, prepend, 4, 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
}
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
if (log_IsKept(LogDEBUG))
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
log_Printf(LogDEBUG, "MP[frag %d]: Send %d bytes on link `%s'\n",
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
mp->out.seq, m_length(m), l->name);
o Increment expected MP fragment numbers correctly, according to SHORTSEQ negotiation. o Don't forget to attach incoming fragments with a number greater than everything else in the queue (rather than leaking memory). o Output the link name with the ``other'' hdlc diagnostic message. o Correct a VJ diagnostic (`COMPPROTO', not `proto').
1998-05-04 03:00:09 +00:00
mp->out.seq = inc_seq(mp->peer_is12bit, mp->out.seq);
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
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
link_PushPacket(l, m, bundle, LINK_QUEUES(l) - 1, PROTO_MP);
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
mp_FillQueues(struct bundle *bundle)
{
struct mp *mp = &bundle->ncp.mp;
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
struct datalink *dl, *fdl;
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
size_t total, add, len;
int thislink, nlinks;
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
u_int32_t begin, end;
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
struct mbuf *m, *mo;
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
thislink = nlinks = 0;
for (fdl = NULL, dl = bundle->links; dl; dl = dl->next) {
o Make sure we adjust our min seq and process any outstanding queued incoming fragments when a link goes down. o Don't use the minimum sequence numbers of links that aren't open. o Understand sequence number wrapping when determining the minimum sequence number. o Add & adjust a few comments.
1998-05-23 17:05:28 +00:00
/* Include non-open links here as mp->out.link will stay more correct */
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
if (!fdl) {
if (thislink == mp->out.link)
fdl = dl;
else
thislink++;
}
nlinks++;
}
if (!fdl) {
fdl = bundle->links;
o Make sure we don't dereference NULL when we've lost all our links. o Use srandomdev() for __FreeBSD__ >= 2, not just >2. o Use srandom((time(NULL)^getpid())+random()), random() when we haven't got srandomdev().
1998-05-06 18:50:12 +00:00
if (!fdl)
return 0;
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
thislink = 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
total = 0;
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
for (dl = fdl; nlinks > 0; dl = dl->next, nlinks--, thislink++) {
if (!dl) {
dl = bundle->links;
thislink = 0;
}
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
if (dl->state != DATALINK_OPEN)
continue;
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
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
if (dl->physical->out)
/* this link has suffered a short write. Let it continue */
continue;
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
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
add = link_QueueLen(&dl->physical->link);
If we're in MP mode with a single open link, MP link level compression isn't open and the links MRU >= our MRRU, send outbound traffic as PROTO_IP rather than PROTO_MP. This shaves some bytes off the front of each packet 'till the second link is brought up. Idea obtained from: Cisco
2000-08-17 14:14:54 +00:00
if (add) {
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
/* this link has got stuff already queued. Let it continue */
If we're in MP mode with a single open link, MP link level compression isn't open and the links MRU >= our MRRU, send outbound traffic as PROTO_IP rather than PROTO_MP. This shaves some bytes off the front of each packet 'till the second link is brought up. Idea obtained from: Cisco
2000-08-17 14:14:54 +00:00
total += add;
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
continue;
If we're in MP mode with a single open link, MP link level compression isn't open and the links MRU >= our MRRU, send outbound traffic as PROTO_IP rather than PROTO_MP. This shaves some bytes off the front of each packet 'till the second link is brought up. Idea obtained from: Cisco
2000-08-17 14:14:54 +00:00
}
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
If we're in MP mode with a single open link, MP link level compression isn't open and the links MRU >= our MRRU, send outbound traffic as PROTO_IP rather than PROTO_MP. This shaves some bytes off the front of each packet 'till the second link is brought up. Idea obtained from: Cisco
2000-08-17 14:14:54 +00:00
if (!link_QueueLen(&mp->link)) {
struct datalink *other;
int mrutoosmall;
/*
* If there's only a single open link in our bundle and we haven't got
* MP level link compression, queue outbound traffic directly via that
* link's protocol stack rather than using the MP link. This results
* in the outbound traffic going out as PROTO_IP rather than PROTO_MP.
*/
for (other = dl->next; other; other = other->next)
if (other->state == DATALINK_OPEN)
break;
mrutoosmall = 0;
if (!other) {
if (dl->physical->link.lcp.his_mru < mp->peer_mrru) {
/*
* Actually, forget it. This test is done against the MRRU rather
* than the packet size so that we don't end up sending some data
* in MP fragments and some data in PROTO_IP packets. That's just
* too likely to upset some ppp implementations.
*/
mrutoosmall = 1;
other = dl;
}
}
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
If we're in MP mode with a single open link, MP link level compression isn't open and the links MRU >= our MRRU, send outbound traffic as PROTO_IP rather than PROTO_MP. This shaves some bytes off the front of each packet 'till the second link is brought up. Idea obtained from: Cisco
2000-08-17 14:14:54 +00:00
if (!ip_PushPacket(other ? &mp->link : &dl->physical->link, bundle))
/* Nothing else to send */
break;
if (mrutoosmall)
log_Printf(LogDEBUG, "Don't send data as PROTO_IP, MRU < MRRU\n");
else if (!other)
log_Printf(LogDEBUG, "Sending data as PROTO_IP, not PROTO_MP\n");
if (!other) {
add = link_QueueLen(&dl->physical->link);
if (add) {
/* this link has got stuff already queued. Let it continue */
total += add;
continue;
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
}
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
}
If we're in MP mode with a single open link, MP link level compression isn't open and the links MRU >= our MRRU, send outbound traffic as PROTO_IP rather than PROTO_MP. This shaves some bytes off the front of each packet 'till the second link is brought up. Idea obtained from: Cisco
2000-08-17 14:14:54 +00:00
}
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
If we're in MP mode with a single open link, MP link level compression isn't open and the links MRU >= our MRRU, send outbound traffic as PROTO_IP rather than PROTO_MP. This shaves some bytes off the front of each packet 'till the second link is brought up. Idea obtained from: Cisco
2000-08-17 14:14:54 +00:00
m = link_Dequeue(&mp->link);
if (m) {
len = m_length(m);
begin = 1;
end = 0;
while (!end) {
if (dl->state == DATALINK_OPEN) {
/* Write at most his_mru bytes to the physical link */
if (len <= dl->physical->link.lcp.his_mru) {
mo = m;
end = 1;
m_settype(mo, MB_MPOUT);
} else {
/* It's > his_mru, chop the packet (`m') into bits */
mo = m_get(dl->physical->link.lcp.his_mru, MB_MPOUT);
len -= mo->m_len;
m = mbuf_Read(m, MBUF_CTOP(mo), mo->m_len);
}
mp_Output(mp, bundle, &dl->physical->link, mo, begin, end);
begin = 0;
}
if (!end) {
nlinks--;
dl = dl->next;
if (!dl) {
dl = bundle->links;
thislink = 0;
} else
thislink++;
}
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
}
}
}
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
mp->out.link = thislink; /* Start here next time */
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
return total;
}
int
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
mp_SetDatalinkBandwidth(struct cmdargs const *arg)
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 val;
Pass all command arguments around with struct cmdargs rather than incrementing argv and decrementing argc. Now individual functions can determine their caller. This also removes the "prefix" hack in FindExec().
1998-04-14 23:17:24 +00:00
if (arg->argc != arg->argn+1)
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
return -1;
If we're in MP mode with a single open link, MP link level compression isn't open and the links MRU >= our MRRU, send outbound traffic as PROTO_IP rather than PROTO_MP. This shaves some bytes off the front of each packet 'till the second link is brought up. Idea obtained from: Cisco
2000-08-17 14:14:54 +00:00
Pass all command arguments around with struct cmdargs rather than incrementing argv and decrementing argc. Now individual functions can determine their caller. This also removes the "prefix" hack in FindExec().
1998-04-14 23:17:24 +00:00
val = atoi(arg->argv[arg->argn]);
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
if (val <= 0) {
log_Printf(LogWARN, "The link bandwidth must be greater than zero\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
return 1;
}
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
arg->cx->mp.bandwidth = val;
if (arg->cx->state == DATALINK_OPEN)
bundle_CalculateBandwidth(arg->bundle);
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
return 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
int
mp_ShowStatus(struct cmdargs const *arg)
{
struct mp *mp = &arg->bundle->ncp.mp;
prompt_Printf(arg->prompt, "Multilink is %sactive\n", mp->active ? "" : "in");
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
if (mp->active) {
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
struct mbuf *m, *lm;
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
int bufs = 0;
Oops, quieten a compiler warning.
1999-06-02 23:06:21 +00:00
lm = NULL;
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
prompt_Printf(arg->prompt, "Socket: %s\n",
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
mp->server.socket.sun_path);
Cosmetic: Make struct mbuf more like kernel mbufs.
1999-12-20 20:29:47 +00:00
for (m = mp->inbufs; m; m = m->m_nextpkt) {
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +00:00
bufs++;
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
lm = m;
}
prompt_Printf(arg->prompt, "Pending frags: %d", bufs);
if (bufs) {
struct mp_header mh;
unsigned long first, last;
first = mp_ReadHeader(mp, mp->inbufs, &mh) ? mh.seq : 0;
last = mp_ReadHeader(mp, lm, &mh) ? mh.seq : 0;
Fix some MP sequence number comparison bogons that are tickled by having different speed links in a bundle. This would manifest itself by having the link occasionally hang, but revive when a new connection is made.... Make ``show mp'' a bit prettier.
1999-06-03 13:29:32 +00:00
prompt_Printf(arg->prompt, " (Have %lu - %lu, want %lu, lowest %lu)\n",
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
first, last, (unsigned long)mp->seq.next_in,
(unsigned long)mp->seq.min_in);
Fix some MP sequence number comparison bogons that are tickled by having different speed links in a bundle. This would manifest itself by having the link occasionally hang, but revive when a new connection is made.... Make ``show mp'' a bit prettier.
1999-06-03 13:29:32 +00:00
prompt_Printf(arg->prompt, " First has %sbegin bit and "
"%send bit", mh.begin ? "" : "no ", mh.end ? "" : "no ");
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
}
prompt_Printf(arg->prompt, "\n");
o Display current link throughput in `show links' (assuming throughput measurement is enabled). o Load balance the links based on weight *and* on a round-robin basis. This makes things fairly even on an output basis. We don't try to allow for the peer sending all his data down one link (and try to send ours back up the other). o Show the number of pending input buffers that can't be processed in ``show mp''. o Fix a typo in the man page.
1998-05-03 22:13:14 +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
prompt_Printf(arg->prompt, "\nMy Side:\n");
if (mp->active) {
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
prompt_Printf(arg->prompt, " Output SEQ: %u\n", mp->out.seq);
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
prompt_Printf(arg->prompt, " MRRU: %u\n", mp->local_mrru);
prompt_Printf(arg->prompt, " Short Seq: %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
mp->local_is12bit ? "on" : "off");
}
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
prompt_Printf(arg->prompt, " Discriminator: %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
mp_Enddisc(mp->cfg.enddisc.class, mp->cfg.enddisc.address,
mp->cfg.enddisc.len));
prompt_Printf(arg->prompt, "\nHis Side:\n");
if (mp->active) {
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
prompt_Printf(arg->prompt, " Auth Name: %s\n", mp->peer.authname);
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
prompt_Printf(arg->prompt, " Input SEQ: %u\n", mp->seq.next_in);
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
prompt_Printf(arg->prompt, " MRRU: %u\n", mp->peer_mrru);
prompt_Printf(arg->prompt, " Short Seq: %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
mp->peer_is12bit ? "on" : "off");
}
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
prompt_Printf(arg->prompt, " Discriminator: %s\n",
mp_Enddisc(mp->peer.enddisc.class, mp->peer.enddisc.address,
mp->peer.enddisc.len));
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
prompt_Printf(arg->prompt, "\nDefaults:\n");
If we're in MP mode with a single open link, MP link level compression isn't open and the links MRU >= our MRRU, send outbound traffic as PROTO_IP rather than PROTO_MP. This shaves some bytes off the front of each packet 'till the second link is brought up. Idea obtained from: Cisco
2000-08-17 14:14:54 +00:00
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
prompt_Printf(arg->prompt, " MRRU: ");
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 (mp->cfg.mrru)
prompt_Printf(arg->prompt, "%d (multilink enabled)\n", mp->cfg.mrru);
else
prompt_Printf(arg->prompt, "disabled\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
prompt_Printf(arg->prompt, " Short Seq: %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
command_ShowNegval(mp->cfg.shortseq));
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
prompt_Printf(arg->prompt, " Discriminator: %s\n",
command_ShowNegval(mp->cfg.negenddisc));
o Obsolete the undocumented ``set weight'' command. o If we're using RADIUS and the RADIUS mtu is less than our peers mru/mrru, reduce our mtu to this value for NetBSD too. o Make struct throughput's sample period dynamic and tweak the ppp version number to reflect the extra stuff being passed through the local domain socket as a result (MP mode). o Measure the current throughput based on the number of samples actually taken rather than on the full sample period. o Keep the throughput statisics persistent while being passed to another ppp invocation through the local domain socket. o When showing throughput statistics after the timer has stopped, use the stopped time for overall calculations, not the current time. Also show the stopped time and how long the current throughput has been sampled for. o Use time() consistently in throughput.c o Tighten up the ``show bundle'' output. o Introduce the ``set bandwidth'' command. o Rewrite the ``set autoload'' command. It now takes three arguments and works based on a rolling bundle throughput average compared against the theoretical bundle bandwidth over a given period (read: it's now functional).
1999-08-05 10:32:16 +00:00
prompt_Printf(arg->prompt, " AutoLoad: min %d%%, max %d%%,"
" period %d secs\n", mp->cfg.autoload.min,
mp->cfg.autoload.max, mp->cfg.autoload.period);
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 0;
}
const char *
mp_Enddisc(u_char c, const char *address, int len)
{
o De-staticise things that don't need to be static. o Bring the static ``ttystate'' into struct prompt so that the tilde context is per prompt and not global. o Comment the remaining static variables so that it's clear why they're static. o Add some XXX comments suggesting that our interface list and our hostname should be re-generated after a signal (say SIGUSR1) so that a machine with PCCARDs has a chance.
1998-06-15 19:06:58 +00:00
static char result[100]; /* Used immediately after it's returned */
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
int f, header;
switch (c) {
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
case ENDDISC_NULL:
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
sprintf(result, "Null Class");
break;
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
case ENDDISC_LOCAL:
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
snprintf(result, sizeof result, "Local Addr: %.*s", len, address);
break;
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
case ENDDISC_IP:
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 (len == 4)
snprintf(result, sizeof result, "IP %s",
inet_ntoa(*(const struct in_addr *)address));
else
sprintf(result, "IP[%d] ???", len);
break;
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
case ENDDISC_MAC:
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 (len == 6) {
const u_char *m = (const u_char *)address;
snprintf(result, sizeof result, "MAC %02x:%02x:%02x:%02x:%02x:%02x",
m[0], m[1], m[2], m[3], m[4], m[5]);
} else
sprintf(result, "MAC[%d] ???", len);
break;
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
case ENDDISC_MAGIC:
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
sprintf(result, "Magic: 0x");
header = strlen(result);
if (len > sizeof result - header - 1)
len = sizeof result - header - 1;
for (f = 0; f < len; f++)
sprintf(result + header + 2 * f, "%02x", address[f]);
break;
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
case ENDDISC_PSN:
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
snprintf(result, sizeof result, "PSN: %.*s", len, address);
break;
default:
sprintf(result, "%d: ", (int)c);
header = strlen(result);
if (len > sizeof result - header - 1)
len = sizeof result - header - 1;
for (f = 0; f < len; f++)
sprintf(result + header + 2 * f, "%02x", address[f]);
break;
}
return result;
}
int
mp_SetEnddisc(struct cmdargs const *arg)
{
struct mp *mp = &arg->bundle->ncp.mp;
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
struct in_addr addr;
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
Loosen our restrictions on setting enddisc, mrru, shortseq, authname and authkey. o Auth{name,key} may additionally be set in PHASE_ESTABLISH. o The others may be set in PHASE_ESTABLISH as long as no links have yet reached DATALINK_LCP.
1998-10-24 01:08:45 +00:00
switch (bundle_Phase(arg->bundle)) {
case PHASE_DEAD:
break;
case PHASE_ESTABLISH:
/* Make sure none of our links are DATALINK_LCP or greater */
if (bundle_HighestState(arg->bundle) >= DATALINK_LCP) {
log_Printf(LogWARN, "enddisc: Only changable before"
" LCP negotiations\n");
return 1;
}
break;
default:
log_Printf(LogWARN, "enddisc: Only changable at phase DEAD/ESTABLISH\n");
return 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
}
if (arg->argc == arg->argn) {
mp->cfg.enddisc.class = 0;
*mp->cfg.enddisc.address = '\0';
mp->cfg.enddisc.len = 0;
Make gcc-2.8.1 build ppp cleanly. Support OpenBSD again.
1998-04-25 10:49:52 +00:00
} else if (arg->argc > arg->argn) {
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
if (!strcasecmp(arg->argv[arg->argn], "label")) {
mp->cfg.enddisc.class = ENDDISC_LOCAL;
strcpy(mp->cfg.enddisc.address, arg->bundle->cfg.label);
mp->cfg.enddisc.len = strlen(mp->cfg.enddisc.address);
} else if (!strcasecmp(arg->argv[arg->argn], "ip")) {
Choose the correct IP number in ``set enddisc {ip,mac}''.
1998-04-24 19:16:20 +00:00
if (arg->bundle->ncp.ipcp.my_ip.s_addr == INADDR_ANY)
addr = arg->bundle->ncp.ipcp.cfg.my_range.ipaddr;
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
else
Choose the correct IP number in ``set enddisc {ip,mac}''.
1998-04-24 19:16:20 +00:00
addr = arg->bundle->ncp.ipcp.my_ip;
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
memcpy(mp->cfg.enddisc.address, &addr.s_addr, sizeof addr.s_addr);
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
mp->cfg.enddisc.class = ENDDISC_IP;
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->cfg.enddisc.len = sizeof arg->bundle->ncp.ipcp.my_ip.s_addr;
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
} else if (!strcasecmp(arg->argv[arg->argn], "mac")) {
struct sockaddr_dl hwaddr;
int s;
Choose the correct IP number in ``set enddisc {ip,mac}''.
1998-04-24 19:16:20 +00:00
if (arg->bundle->ncp.ipcp.my_ip.s_addr == INADDR_ANY)
addr = arg->bundle->ncp.ipcp.cfg.my_range.ipaddr;
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
else
Choose the correct IP number in ``set enddisc {ip,mac}''.
1998-04-24 19:16:20 +00:00
addr = arg->bundle->ncp.ipcp.my_ip;
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
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
s = ID0socket(AF_INET, SOCK_DGRAM, 0);
if (s < 0) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogERROR, "set enddisc: socket(): %s\n", strerror(errno));
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
return 2;
}
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
if (get_ether_addr(s, addr, &hwaddr)) {
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
mp->cfg.enddisc.class = ENDDISC_MAC;
memcpy(mp->cfg.enddisc.address, hwaddr.sdl_data + hwaddr.sdl_nlen,
hwaddr.sdl_alen);
mp->cfg.enddisc.len = hwaddr.sdl_alen;
} else {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogWARN, "set enddisc: Can't locate MAC address for %s\n",
Choose the correct IP number in ``set enddisc {ip,mac}''.
1998-04-24 19:16:20 +00:00
inet_ntoa(addr));
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
close(s);
return 4;
}
close(s);
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
} else if (!strcasecmp(arg->argv[arg->argn], "magic")) {
int f;
randinit();
for (f = 0; f < 20; f += sizeof(long))
*(long *)(mp->cfg.enddisc.address + f) = random();
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
mp->cfg.enddisc.class = ENDDISC_MAGIC;
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->cfg.enddisc.len = 20;
} else if (!strcasecmp(arg->argv[arg->argn], "psn")) {
if (arg->argc > arg->argn+1) {
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
mp->cfg.enddisc.class = ENDDISC_PSN;
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
strcpy(mp->cfg.enddisc.address, arg->argv[arg->argn+1]);
mp->cfg.enddisc.len = strlen(mp->cfg.enddisc.address);
} else {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogWARN, "PSN endpoint requires additional data\n");
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
return 5;
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
}
} else {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogWARN, "%s: Unrecognised endpoint type\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
arg->argv[arg->argn]);
Create #defines for all endpoint discriptor classes, and allow `set enddisc mac'.
1998-04-23 23:50:40 +00:00
return 6;
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
}
Make gcc-2.8.1 build ppp cleanly. Support OpenBSD again.
1998-04-25 10:49:52 +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
return 0;
}
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
static int
To avoid namespace polution in NetBSD: ``struct descriptor'' -> ``struct fdescriptor''
2000-03-14 01:46:09 +00:00
mpserver_UpdateSet(struct fdescriptor *d, fd_set *r, fd_set *w, fd_set *e,
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
int *n)
{
struct mpserver *s = descriptor2mpserver(d);
o Move our prompt descriptor list outside of the bundle. It's now dealt with by the `server' object. This simplifies things as we only have one list of prompt descriptors and the log_ routines check prompt::logactive to determine whether it should be used for output. o Include the MP socket UpdateSet() result in bundle::UpdateSet(). o Don't select on the tun device unless we're in NETWORK phase or AUTO mode. o Stop the idle timer when we go to DEAD phase. We may have transferred a link and not had a chance to kill it. o Don't fail when trying to unlink our transferred datalink from our descriptor lists just before the transfer. o Add our link descriptor to the write set if we got a short write the last time (physical::out is set). o Log the connection source address when a connection is closed. o Remove descriptor::next field. Descriptor lists are not required any more.
1998-05-23 22:24:50 +00:00
int result;
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
o Move our prompt descriptor list outside of the bundle. It's now dealt with by the `server' object. This simplifies things as we only have one list of prompt descriptors and the log_ routines check prompt::logactive to determine whether it should be used for output. o Include the MP socket UpdateSet() result in bundle::UpdateSet(). o Don't select on the tun device unless we're in NETWORK phase or AUTO mode. o Stop the idle timer when we go to DEAD phase. We may have transferred a link and not had a chance to kill it. o Don't fail when trying to unlink our transferred datalink from our descriptor lists just before the transfer. o Add our link descriptor to the write set if we got a short write the last time (physical::out is set). o Log the connection source address when a connection is closed. o Remove descriptor::next field. Descriptor lists are not required any more.
1998-05-23 22:24:50 +00:00
result = 0;
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
if (s->send.dl != NULL) {
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
/* We've connect()ed */
o Rename datalinks as soon as the name has been received so that LQM and HDLC timer diagnostics come out with the correct name. o Don't send an LQR immediately upon reviving a datalink. Leave it 'till the next timeout. o Add the link name to some more LQR diagnostics. o Break out of the main loop when a descriptor exception is seen in select(). o Remove the evil nointr_[u]sleep() functions. Timers should be (and are) used instead. o Treat a read() of 0 bytes as an error that's fatal to the link on which the read() is done. We should never read() 0 after select() says there's something there - not unless the link has been closed by the other side. o Write the data seen before a HDLC header to the terminal in `term' mode, *not* back to the modem :-/ o Initialise our transmitted file descriptor before starting any timers. o Only send data links that have *no* pending output data. This means that our final ACK will be written rather than being nuked with the datalink transmission.
1998-05-08 01:15:19 +00:00
if (!link_QueueLen(&s->send.dl->physical->link) &&
!s->send.dl->physical->out) {
/* Only send if we've transmitted all our data (i.e. the ConfigAck) */
o Move our prompt descriptor list outside of the bundle. It's now dealt with by the `server' object. This simplifies things as we only have one list of prompt descriptors and the log_ routines check prompt::logactive to determine whether it should be used for output. o Include the MP socket UpdateSet() result in bundle::UpdateSet(). o Don't select on the tun device unless we're in NETWORK phase or AUTO mode. o Stop the idle timer when we go to DEAD phase. We may have transferred a link and not had a chance to kill it. o Don't fail when trying to unlink our transferred datalink from our descriptor lists just before the transfer. o Add our link descriptor to the write set if we got a short write the last time (physical::out is set). o Log the connection source address when a connection is closed. o Remove descriptor::next field. Descriptor lists are not required any more.
1998-05-23 22:24:50 +00:00
result -= datalink_RemoveFromSet(s->send.dl, r, w, e);
o Rename datalinks as soon as the name has been received so that LQM and HDLC timer diagnostics come out with the correct name. o Don't send an LQR immediately upon reviving a datalink. Leave it 'till the next timeout. o Add the link name to some more LQR diagnostics. o Break out of the main loop when a descriptor exception is seen in select(). o Remove the evil nointr_[u]sleep() functions. Timers should be (and are) used instead. o Treat a read() of 0 bytes as an error that's fatal to the link on which the read() is done. We should never read() 0 after select() says there's something there - not unless the link has been closed by the other side. o Write the data seen before a HDLC header to the terminal in `term' mode, *not* back to the modem :-/ o Initialise our transmitted file descriptor before starting any timers. o Only send data links that have *no* pending output data. This means that our final ACK will be written rather than being nuked with the datalink transmission.
1998-05-08 01:15:19 +00:00
bundle_SendDatalink(s->send.dl, s->fd, &s->socket);
s->send.dl = NULL;
s->fd = -1;
o Call bundle_LinkClosed() when transferring a datalink so that the bundle has the opportunity to go PHASE_DEAD and cleanup the interface (if it's the last link). o Regnerate our phys_type value when we transfer the link. o Always clean up the interface when destroying our bundle in case we're abending. o Always clean up our interface when the last link is gone rather than delaying things 'till exit time in the -direct case (the interface is useless anyway). Do this *after* slamming down our NCPs (if they're still around). o Our MP server descriptor now clears the relevent device descriptor from our descriptor [fd]sets when a datalink is on death-row (to be transferred to another running ppp), thus avoiding the possibility of passing a bum descriptor to select() and having ppp abend. o Handle the MP socket descriptor functions from within the bundle descriptor functions. Now we ensure that the MP socket descriptor functions see the descriptor sets *after* they've been seen by our datalinks. o Add/fix a few more comments.
1998-05-15 18:21:45 +00:00
} else
/* Never read from a datalink that's on death row ! */
o Move our prompt descriptor list outside of the bundle. It's now dealt with by the `server' object. This simplifies things as we only have one list of prompt descriptors and the log_ routines check prompt::logactive to determine whether it should be used for output. o Include the MP socket UpdateSet() result in bundle::UpdateSet(). o Don't select on the tun device unless we're in NETWORK phase or AUTO mode. o Stop the idle timer when we go to DEAD phase. We may have transferred a link and not had a chance to kill it. o Don't fail when trying to unlink our transferred datalink from our descriptor lists just before the transfer. o Add our link descriptor to the write set if we got a short write the last time (physical::out is set). o Log the connection source address when a connection is closed. o Remove descriptor::next field. Descriptor lists are not required any more.
1998-05-23 22:24:50 +00:00
result -= datalink_RemoveFromSet(s->send.dl, r, NULL, NULL);
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
} else if (r && s->fd >= 0) {
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
if (*n < s->fd + 1)
*n = s->fd + 1;
FD_SET(s->fd, r);
o Protect against expected NULL fdset pointers. o Log FD_SET()s in LogTIMER. o Identify the descriptor that causes an EBADF from select() if LogTIMER is enabled (then exit). o Call the MP server UpdateSet() function after calling the UpdateSet() for all links - the link may enter PHASE_TERMINATE and bring down the MP server - breaking the imminent select().
1998-05-10 22:20:20 +00:00
log_Printf(LogTIMER, "mp: fdset(r) %d\n", s->fd);
o Move our prompt descriptor list outside of the bundle. It's now dealt with by the `server' object. This simplifies things as we only have one list of prompt descriptors and the log_ routines check prompt::logactive to determine whether it should be used for output. o Include the MP socket UpdateSet() result in bundle::UpdateSet(). o Don't select on the tun device unless we're in NETWORK phase or AUTO mode. o Stop the idle timer when we go to DEAD phase. We may have transferred a link and not had a chance to kill it. o Don't fail when trying to unlink our transferred datalink from our descriptor lists just before the transfer. o Add our link descriptor to the write set if we got a short write the last time (physical::out is set). o Log the connection source address when a connection is closed. o Remove descriptor::next field. Descriptor lists are not required any more.
1998-05-23 22:24:50 +00:00
result++;
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
}
o Move our prompt descriptor list outside of the bundle. It's now dealt with by the `server' object. This simplifies things as we only have one list of prompt descriptors and the log_ routines check prompt::logactive to determine whether it should be used for output. o Include the MP socket UpdateSet() result in bundle::UpdateSet(). o Don't select on the tun device unless we're in NETWORK phase or AUTO mode. o Stop the idle timer when we go to DEAD phase. We may have transferred a link and not had a chance to kill it. o Don't fail when trying to unlink our transferred datalink from our descriptor lists just before the transfer. o Add our link descriptor to the write set if we got a short write the last time (physical::out is set). o Log the connection source address when a connection is closed. o Remove descriptor::next field. Descriptor lists are not required any more.
1998-05-23 22:24:50 +00:00
return result;
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
}
static int
To avoid namespace polution in NetBSD: ``struct descriptor'' -> ``struct fdescriptor''
2000-03-14 01:46:09 +00:00
mpserver_IsSet(struct fdescriptor *d, const fd_set *fdset)
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
{
struct mpserver *s = descriptor2mpserver(d);
return s->fd >= 0 && FD_ISSET(s->fd, fdset);
}
static void
To avoid namespace polution in NetBSD: ``struct descriptor'' -> ``struct fdescriptor''
2000-03-14 01:46:09 +00:00
mpserver_Read(struct fdescriptor *d, struct bundle *bundle, const fd_set *fdset)
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
{
struct mpserver *s = descriptor2mpserver(d);
Rewrite the link descriptor transfer code in MP mode. Previously, ppp attempted to bind() to a local domain tcp socket based on the peer authname & enddisc. If it succeeded, it listen()ed and became MP server. If it failed, it connect()ed and became MP client. The server then select()ed on the descriptor, accept()ed it and wrote its pid to it then read the link data & link file descriptor, and finally sent an ack (``!''). The client would read() the server pid, transfer the link lock to that pid, send the link data & descriptor and read the ack. It would then close the descriptor and clean up. There was a race between the bind() and listen() where someone could attempt to connect() and fail. This change removes the race. Now ppp makes the RCVBUF big enough on a socket descriptor and attempts to bind() to a local domain *udp* socket (same name as before). If it succeeds, it becomes MP server. If it fails, it sets the SNDBUF and connect()s, becoming MP client. The server select()s on the descriptor and recvmsg()s the message, insisting on at least two descriptors (plus the link data). It uses the second descriptor to write() its pid then read()s an ack (``!''). The client creates a socketpair() and sendmsg()s the link data, link descriptor and one of the socketpair descriptors. It then read()s the server pid from the other socketpair descriptor, transfers any locks and write()s an ack. Now, there can be no race, and a connect() failure indicates a stale socket file. This also fixes MP ppp over ethernet, where the struct msghdr was being misconstructed when transferring the control socket descriptor. Also, if we fail to send the link, don't hang around in a ``session owner'' state, just do the setsid() and fork() if it's required to disown a tty. UDP idea suggested by: Chris Bennet from Mindspring at FreeBSDCon
1999-11-25 02:47:04 +00:00
bundle_ReceiveDatalink(bundle, s->fd);
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
}
o If we come out of select() with only write descriptors that end up writing zero bytes, sleep for 1/10 of a second so that we don't end up using up too much cpu. This should only ever happen on systems that wrongly report a descriptor as writable despite the tty buffer being full. Discussed with: Jeff Evarts o Do an initial run-time check to see if select() alters the passed timeval. This knowledge isn't yet used, but will be soon.
1998-06-24 19:33:36 +00:00
static int
Various whitespace changes. Make some functions static.
2000-10-30 00:15:29 +00:00
mpserver_Write(struct fdescriptor *d, struct bundle *bundle,
const fd_set *fdset)
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
{
/* We never want to write here ! */
Change some log levels. ALERTs are only logged when something that can't happen happens or when everyone needs to know. ERRORs are only logged when something unexpected happens.
1998-06-16 19:40:42 +00:00
log_Printf(LogALERT, "mpserver_Write: Internal error: Bad call !\n");
o If we come out of select() with only write descriptors that end up writing zero bytes, sleep for 1/10 of a second so that we don't end up using up too much cpu. This should only ever happen on systems that wrongly report a descriptor as writable despite the tty buffer being full. Discussed with: Jeff Evarts o Do an initial run-time check to see if select() alters the passed timeval. This knowledge isn't yet used, but will be soon.
1998-06-24 19:33:36 +00:00
return 0;
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
}
void
mpserver_Init(struct mpserver *s)
{
s->desc.type = MPSERVER_DESCRIPTOR;
s->desc.UpdateSet = mpserver_UpdateSet;
s->desc.IsSet = mpserver_IsSet;
s->desc.Read = mpserver_Read;
s->desc.Write = mpserver_Write;
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
s->send.dl = NULL;
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
s->fd = -1;
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
memset(&s->socket, '\0', sizeof s->socket);
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
}
int
mpserver_Open(struct mpserver *s, struct peerid *peer)
{
Change the way we transfer links (again). The previous method avoided all race conditions, but suffered from sometimes running out of buffer space if enough clients were piled up at the same time. Now, the client pushes the link descriptor, one end of a socketpair() and the ppp version via sendmsg() at the server. The server replies with a pid. The client then transfers any link lock with uu_lock_txfr() and writev()s the actual link contents. The socketpair is now the only place we need to have large socket buffers and the bind()ed socket can keep the default 4k buffer while still handling around 90 racing clients.
1999-11-30 23:52:37 +00:00
int f, l;
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
mode_t mask;
if (s->fd != -1) {
Change some log levels. ALERTs are only logged when something that can't happen happens or when everyone needs to know. ERRORs are only logged when something unexpected happens.
1998-06-16 19:40:42 +00:00
log_Printf(LogALERT, "Internal error ! mpserver already open\n");
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
mpserver_Close(s);
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
}
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
l = snprintf(s->socket.sun_path, sizeof s->socket.sun_path, "%sppp-%s-%02x-",
_PATH_VARRUN, peer->authname, peer->enddisc.class);
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
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
for (f = 0; f < peer->enddisc.len && l < sizeof s->socket.sun_path - 2; f++) {
snprintf(s->socket.sun_path + l, sizeof s->socket.sun_path - l,
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
"%02x", *(u_char *)(peer->enddisc.address+f));
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
l += 2;
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
}
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
s->socket.sun_family = AF_LOCAL;
s->socket.sun_len = sizeof s->socket;
Rewrite the link descriptor transfer code in MP mode. Previously, ppp attempted to bind() to a local domain tcp socket based on the peer authname & enddisc. If it succeeded, it listen()ed and became MP server. If it failed, it connect()ed and became MP client. The server then select()ed on the descriptor, accept()ed it and wrote its pid to it then read the link data & link file descriptor, and finally sent an ack (``!''). The client would read() the server pid, transfer the link lock to that pid, send the link data & descriptor and read the ack. It would then close the descriptor and clean up. There was a race between the bind() and listen() where someone could attempt to connect() and fail. This change removes the race. Now ppp makes the RCVBUF big enough on a socket descriptor and attempts to bind() to a local domain *udp* socket (same name as before). If it succeeds, it becomes MP server. If it fails, it sets the SNDBUF and connect()s, becoming MP client. The server select()s on the descriptor and recvmsg()s the message, insisting on at least two descriptors (plus the link data). It uses the second descriptor to write() its pid then read()s an ack (``!''). The client creates a socketpair() and sendmsg()s the link data, link descriptor and one of the socketpair descriptors. It then read()s the server pid from the other socketpair descriptor, transfers any locks and write()s an ack. Now, there can be no race, and a connect() failure indicates a stale socket file. This also fixes MP ppp over ethernet, where the struct msghdr was being misconstructed when transferring the control socket descriptor. Also, if we fail to send the link, don't hang around in a ``session owner'' state, just do the setsid() and fork() if it's required to disown a tty. UDP idea suggested by: Chris Bennet from Mindspring at FreeBSDCon
1999-11-25 02:47:04 +00:00
s->fd = ID0socket(PF_LOCAL, SOCK_DGRAM, 0);
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
if (s->fd < 0) {
Rewrite the link descriptor transfer code in MP mode. Previously, ppp attempted to bind() to a local domain tcp socket based on the peer authname & enddisc. If it succeeded, it listen()ed and became MP server. If it failed, it connect()ed and became MP client. The server then select()ed on the descriptor, accept()ed it and wrote its pid to it then read the link data & link file descriptor, and finally sent an ack (``!''). The client would read() the server pid, transfer the link lock to that pid, send the link data & descriptor and read the ack. It would then close the descriptor and clean up. There was a race between the bind() and listen() where someone could attempt to connect() and fail. This change removes the race. Now ppp makes the RCVBUF big enough on a socket descriptor and attempts to bind() to a local domain *udp* socket (same name as before). If it succeeds, it becomes MP server. If it fails, it sets the SNDBUF and connect()s, becoming MP client. The server select()s on the descriptor and recvmsg()s the message, insisting on at least two descriptors (plus the link data). It uses the second descriptor to write() its pid then read()s an ack (``!''). The client creates a socketpair() and sendmsg()s the link data, link descriptor and one of the socketpair descriptors. It then read()s the server pid from the other socketpair descriptor, transfers any locks and write()s an ack. Now, there can be no race, and a connect() failure indicates a stale socket file. This also fixes MP ppp over ethernet, where the struct msghdr was being misconstructed when transferring the control socket descriptor. Also, if we fail to send the link, don't hang around in a ``session owner'' state, just do the setsid() and fork() if it's required to disown a tty. UDP idea suggested by: Chris Bennet from Mindspring at FreeBSDCon
1999-11-25 02:47:04 +00:00
log_Printf(LogERROR, "mpserver: socket(): %s\n", strerror(errno));
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
return MPSERVER_FAILED;
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
}
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
setsockopt(s->fd, SOL_SOCKET, SO_REUSEADDR, (struct sockaddr *)&s->socket,
sizeof s->socket);
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
mask = umask(0177);
Rewrite the link descriptor transfer code in MP mode. Previously, ppp attempted to bind() to a local domain tcp socket based on the peer authname & enddisc. If it succeeded, it listen()ed and became MP server. If it failed, it connect()ed and became MP client. The server then select()ed on the descriptor, accept()ed it and wrote its pid to it then read the link data & link file descriptor, and finally sent an ack (``!''). The client would read() the server pid, transfer the link lock to that pid, send the link data & descriptor and read the ack. It would then close the descriptor and clean up. There was a race between the bind() and listen() where someone could attempt to connect() and fail. This change removes the race. Now ppp makes the RCVBUF big enough on a socket descriptor and attempts to bind() to a local domain *udp* socket (same name as before). If it succeeds, it becomes MP server. If it fails, it sets the SNDBUF and connect()s, becoming MP client. The server select()s on the descriptor and recvmsg()s the message, insisting on at least two descriptors (plus the link data). It uses the second descriptor to write() its pid then read()s an ack (``!''). The client creates a socketpair() and sendmsg()s the link data, link descriptor and one of the socketpair descriptors. It then read()s the server pid from the other socketpair descriptor, transfers any locks and write()s an ack. Now, there can be no race, and a connect() failure indicates a stale socket file. This also fixes MP ppp over ethernet, where the struct msghdr was being misconstructed when transferring the control socket descriptor. Also, if we fail to send the link, don't hang around in a ``session owner'' state, just do the setsid() and fork() if it's required to disown a tty. UDP idea suggested by: Chris Bennet from Mindspring at FreeBSDCon
1999-11-25 02:47:04 +00:00
/*
Change the way we transfer links (again). The previous method avoided all race conditions, but suffered from sometimes running out of buffer space if enough clients were piled up at the same time. Now, the client pushes the link descriptor, one end of a socketpair() and the ppp version via sendmsg() at the server. The server replies with a pid. The client then transfers any link lock with uu_lock_txfr() and writev()s the actual link contents. The socketpair is now the only place we need to have large socket buffers and the bind()ed socket can keep the default 4k buffer while still handling around 90 racing clients.
1999-11-30 23:52:37 +00:00
* Try to bind the socket. If we succeed we play server, if we fail
* we connect() and hand the link off.
Rewrite the link descriptor transfer code in MP mode. Previously, ppp attempted to bind() to a local domain tcp socket based on the peer authname & enddisc. If it succeeded, it listen()ed and became MP server. If it failed, it connect()ed and became MP client. The server then select()ed on the descriptor, accept()ed it and wrote its pid to it then read the link data & link file descriptor, and finally sent an ack (``!''). The client would read() the server pid, transfer the link lock to that pid, send the link data & descriptor and read the ack. It would then close the descriptor and clean up. There was a race between the bind() and listen() where someone could attempt to connect() and fail. This change removes the race. Now ppp makes the RCVBUF big enough on a socket descriptor and attempts to bind() to a local domain *udp* socket (same name as before). If it succeeds, it becomes MP server. If it fails, it sets the SNDBUF and connect()s, becoming MP client. The server select()s on the descriptor and recvmsg()s the message, insisting on at least two descriptors (plus the link data). It uses the second descriptor to write() its pid then read()s an ack (``!''). The client creates a socketpair() and sendmsg()s the link data, link descriptor and one of the socketpair descriptors. It then read()s the server pid from the other socketpair descriptor, transfers any locks and write()s an ack. Now, there can be no race, and a connect() failure indicates a stale socket file. This also fixes MP ppp over ethernet, where the struct msghdr was being misconstructed when transferring the control socket descriptor. Also, if we fail to send the link, don't hang around in a ``session owner'' state, just do the setsid() and fork() if it's required to disown a tty. UDP idea suggested by: Chris Bennet from Mindspring at FreeBSDCon
1999-11-25 02:47:04 +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
if (ID0bind_un(s->fd, &s->socket) < 0) {
if (errno != EADDRINUSE) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogPHASE, "mpserver: can't create bundle socket %s (%s)\n",
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
s->socket.sun_path, strerror(errno));
umask(mask);
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
close(s->fd);
s->fd = -1;
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
return MPSERVER_FAILED;
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
}
Rewrite the link descriptor transfer code in MP mode. Previously, ppp attempted to bind() to a local domain tcp socket based on the peer authname & enddisc. If it succeeded, it listen()ed and became MP server. If it failed, it connect()ed and became MP client. The server then select()ed on the descriptor, accept()ed it and wrote its pid to it then read the link data & link file descriptor, and finally sent an ack (``!''). The client would read() the server pid, transfer the link lock to that pid, send the link data & descriptor and read the ack. It would then close the descriptor and clean up. There was a race between the bind() and listen() where someone could attempt to connect() and fail. This change removes the race. Now ppp makes the RCVBUF big enough on a socket descriptor and attempts to bind() to a local domain *udp* socket (same name as before). If it succeeds, it becomes MP server. If it fails, it sets the SNDBUF and connect()s, becoming MP client. The server select()s on the descriptor and recvmsg()s the message, insisting on at least two descriptors (plus the link data). It uses the second descriptor to write() its pid then read()s an ack (``!''). The client creates a socketpair() and sendmsg()s the link data, link descriptor and one of the socketpair descriptors. It then read()s the server pid from the other socketpair descriptor, transfers any locks and write()s an ack. Now, there can be no race, and a connect() failure indicates a stale socket file. This also fixes MP ppp over ethernet, where the struct msghdr was being misconstructed when transferring the control socket descriptor. Also, if we fail to send the link, don't hang around in a ``session owner'' state, just do the setsid() and fork() if it's required to disown a tty. UDP idea suggested by: Chris Bennet from Mindspring at FreeBSDCon
1999-11-25 02:47:04 +00:00
Change the way we transfer links (again). The previous method avoided all race conditions, but suffered from sometimes running out of buffer space if enough clients were piled up at the same time. Now, the client pushes the link descriptor, one end of a socketpair() and the ppp version via sendmsg() at the server. The server replies with a pid. The client then transfers any link lock with uu_lock_txfr() and writev()s the actual link contents. The socketpair is now the only place we need to have large socket buffers and the bind()ed socket can keep the default 4k buffer while still handling around 90 racing clients.
1999-11-30 23:52:37 +00:00
/* So we're the sender */
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
umask(mask);
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
if (ID0connect_un(s->fd, &s->socket) < 0) {
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogPHASE, "mpserver: can't connect to bundle socket %s (%s)\n",
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
s->socket.sun_path, strerror(errno));
if (errno == ECONNREFUSED)
Rewrite the link descriptor transfer code in MP mode. Previously, ppp attempted to bind() to a local domain tcp socket based on the peer authname & enddisc. If it succeeded, it listen()ed and became MP server. If it failed, it connect()ed and became MP client. The server then select()ed on the descriptor, accept()ed it and wrote its pid to it then read the link data & link file descriptor, and finally sent an ack (``!''). The client would read() the server pid, transfer the link lock to that pid, send the link data & descriptor and read the ack. It would then close the descriptor and clean up. There was a race between the bind() and listen() where someone could attempt to connect() and fail. This change removes the race. Now ppp makes the RCVBUF big enough on a socket descriptor and attempts to bind() to a local domain *udp* socket (same name as before). If it succeeds, it becomes MP server. If it fails, it sets the SNDBUF and connect()s, becoming MP client. The server select()s on the descriptor and recvmsg()s the message, insisting on at least two descriptors (plus the link data). It uses the second descriptor to write() its pid then read()s an ack (``!''). The client creates a socketpair() and sendmsg()s the link data, link descriptor and one of the socketpair descriptors. It then read()s the server pid from the other socketpair descriptor, transfers any locks and write()s an ack. Now, there can be no race, and a connect() failure indicates a stale socket file. This also fixes MP ppp over ethernet, where the struct msghdr was being misconstructed when transferring the control socket descriptor. Also, if we fail to send the link, don't hang around in a ``session owner'' state, just do the setsid() and fork() if it's required to disown a tty. UDP idea suggested by: Chris Bennet from Mindspring at FreeBSDCon
1999-11-25 02:47:04 +00:00
log_Printf(LogPHASE, " The previous server died badly !\n");
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
close(s->fd);
s->fd = -1;
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
return MPSERVER_FAILED;
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
}
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
/* Donate our link to the other guy */
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
return MPSERVER_CONNECTED;
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
}
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
return MPSERVER_LISTENING;
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
}
void
mpserver_Close(struct mpserver *s)
{
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
if (s->send.dl != NULL) {
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
bundle_SendDatalink(s->send.dl, s->fd, &s->socket);
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
s->send.dl = NULL;
o Transfer file descriptors using {send,recv}msg & SCM_RIGHTS, using the scatter/gather array to transfer the link information. The whole link is now passed in one message. This is far better than the two `/bin/cat' processes per additional link :-) I remember years ago thinking that file descriptor transferral would be a really amazing facility ! Suggested by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> and: Eivind Eklund <eivind@yes.no>
1998-05-02 21:57:50 +00:00
s->fd = -1;
} else if (s->fd >= 0) {
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
close(s->fd);
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
if (ID0unlink(s->socket.sun_path) == -1)
Cosmetic: Make our external function names consistent.
1998-05-01 19:26:12 +00:00
log_Printf(LogERROR, "%s: Failed to remove: %s\n", s->socket.sun_path,
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
strerror(errno));
memset(&s->socket, '\0', sizeof s->socket);
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
s->fd = -1;
}
}
o Make sure we adjust our min seq and process any outstanding queued incoming fragments when a link goes down. o Don't use the minimum sequence numbers of links that aren't open. o Understand sequence number wrapping when determining the minimum sequence number. o Add & adjust a few comments.
1998-05-23 17:05:28 +00:00
void
mp_LinkLost(struct mp *mp, struct datalink *dl)
{
if (mp->seq.min_in == dl->mp.seq)
/* We've lost the link that's holding everything up ! */
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
mp_Assemble(mp, NULL, NULL);
o Make sure we adjust our min seq and process any outstanding queued incoming fragments when a link goes down. o Don't use the minimum sequence numbers of links that aren't open. o Understand sequence number wrapping when determining the minimum sequence number. o Add & adjust a few comments.
1998-05-23 17:05:28 +00:00
}
If we've got a full output buffer queue and cannot send anything for two mintues (see ``set choked'' and ``show bundle''), nuke the ip, mp and link level buffer queues. This should fix problems where ``ppp -auto'' seems to stop responding after failing to connect to the peer a few times.
1998-08-25 17:48:43 +00:00
void
mp_DeleteQueue(struct mp *mp)
{
link_DeleteQueue(&mp->link);
}
Reference in New Issue Copy Permalink