]> git.ozlabs.org Git - ppp.git/blobdiff - pppd/pppd.8
Remove the requirement that redistributions in binary form reproduce
[ppp.git] / pppd / pppd.8
index 229708a12b1cfa5947cb6af24e00cae18f283a73..0ccb2e602367a311cab622373abb609c555d1bc2 100644 (file)
@@ -1,62 +1,77 @@
 .\" manual page [] for pppd 2.4
-.\" $Id: pppd.8,v 1.60 2002/03/01 14:39:18 dfs Exp $
+.\" $Id: pppd.8,v 1.79 2004/11/04 10:02:26 paulus Exp $
 .\" SH section heading
 .\" SS subsection heading
 .\" LP paragraph
 .\" IP indented paragraph
 .\" TP hanging label
+.\" 
+.\" Copyright (c) 1993-2003 Paul Mackerras <paulus@samba.org>
+.\"
+.\" Permission to use, copy, modify, and distribute this software for any
+.\" purpose with or without fee is hereby granted, provided that the above
+.\" copyright notice and this permission notice appear in all copies.
+.\"
+.\" THIS SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
+.\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+.\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+.\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+.\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+.\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+.\"
 .TH PPPD 8
 .SH NAME
-pppd \- Point to Point Protocol daemon
+pppd \- Point-to-Point Protocol Daemon
 .SH SYNOPSIS
 .B pppd
 [
-.I tty_name
-] [
-.I speed
-] [
 .I options
 ]
 .SH DESCRIPTION
 .LP
-The Point-to-Point Protocol (PPP) provides a method for transmitting
-datagrams over serial point-to-point links.  PPP
-is composed of three parts: a method for encapsulating datagrams over
-serial links, an extensible Link Control Protocol (LCP), and
-a family of Network Control Protocols (NCP) for establishing
-and configuring different network-layer protocols.
-.LP
-The encapsulation scheme is provided by driver code in the kernel.
-Pppd provides the basic LCP, authentication support, and an NCP for
-establishing and configuring the Internet Protocol (IP) (called the IP
-Control Protocol, IPCP).
+PPP is the protocol used for establishing internet links over dial-up
+modems, DSL connections, and many other types of point-to-point
+links.  The \fIpppd\fR daemon works together with the kernel PPP
+driver to establish and maintain a PPP link with another system
+(called the \fIpeer\fR) and to negotiate Internet Protocol (IP)
+addresses for each end of the link.  Pppd can also authenticate the
+peer and/or supply authentication information to the peer.  PPP can be
+used with other network protocols besides IP, but such use is becoming
+increasingly rare.
 .SH FREQUENTLY USED OPTIONS
 .TP
-.I <tty_name>
-Communicate over the named device.  The string "/dev/" is prepended if
-necessary.  If no device name is given, or if the name of the terminal
+.I ttyname
+Use the serial port called \fIttyname\fR to communicate with the
+peer.  The string "/dev/" is prepended to \fIttyname\fR to form the
+name of the device to open.  If no device name is given, or if the
+name of the terminal
 connected to the standard input is given, pppd will use that terminal,
 and will not fork to put itself in the background.  A value for this
 option from a privileged source cannot be overridden by a
 non-privileged user.
 .TP
-.I <speed>
-Set the baud rate to <speed> (a decimal number).  On systems such as
+.I speed
+An option that is a decimal number is taken as the desired baud rate
+for the serial device.  On systems such as
 4.4BSD and NetBSD, any speed can be specified.  Other systems
-(e.g. SunOS) allow only a limited set of speeds.
-.TP
-.B asyncmap \fI<map>
-Set the async character map to <map>.  This map describes which
-control characters cannot be successfully received over the serial
-line.  Pppd will ask the peer to send these characters as a 2-byte
-escape sequence.  The argument is a 32 bit hex number with each bit
-representing a character to escape.  Bit 0 (00000001) represents the
-character 0x00; bit 31 (80000000) represents the character 0x1f or ^_.
+(e.g. Linux, SunOS) only support the commonly-used baud rates.
+.TP
+.B asyncmap \fImap
+This option sets the Async-Control-Character-Map (ACCM) for this end
+of the link.  The ACCM is a set of 32 bits, one for each of the
+ASCII control characters with values from 0 to 31, where a 1 bit
+indicates that the corresponding control character should not be used
+in PPP packets sent to this system.  The map is encoded as a
+hexadecimal number (without a leading 0x) where the least significant
+bit (00000001) represents character 0 and the most significant bit
+(80000000) represents character 31.
+Pppd will ask the peer to send these characters as a 2-byte
+escape sequence.
 If multiple \fIasyncmap\fR options are given, the values are ORed
-together.  If no \fIasyncmap\fR option is given, no async character
-map will be negotiated for the receive direction; the peer should then
-escape \fIall\fR control characters.  To escape transmitted
-characters, use the \fIescape\fR option.
+together.  If no \fIasyncmap\fR option is given, the default is zero,
+so pppd will ask the peer not to escape any control characters.
+To escape transmitted characters, use the \fIescape\fR option.
 .TP
 .B auth
 Require the peer to authenticate itself before allowing network
@@ -73,15 +88,21 @@ or include .. as a pathname component.  The format of the options file
 is described below.
 .TP
 .B connect \fIscript
-Use the executable or shell command specified by \fIscript\fR to set
-up the serial line.  This script would typically use the chat(8)
-program to dial the modem and start the remote ppp session.  A value
+Usually there is something which needs to be done to prepare the link
+before the PPP protocol can be started; for instance, with a dial-up
+modem, commands need to be sent to the modem to dial the appropriate
+phone number.  This option specifies an command for pppd to execute
+(by passing it to a shell) before attempting to start PPP negotiation.
+The chat (8) program is often useful here, as it provides a way to
+send arbitrary strings to a modem and respond to received characters.
+A value
 for this option from a privileged source cannot be overridden by a
 non-privileged user.
 .TP
 .B crtscts
-Use hardware flow control (i.e. RTS/CTS) to control the flow of
-data on the serial port.  If neither the \fIcrtscts\fR, the
+Specifies that pppd should set the serial port to use hardware flow
+control using the RTS and CTS signals in the RS-232 interface.
+If neither the \fIcrtscts\fR, the
 \fInocrtscts\fR, the \fIcdtrcts\fR nor the \fInocdtrcts\fR option
 is given, the hardware flow control setting for the serial port is
 left unchanged.
@@ -89,7 +110,7 @@ Some serial ports (such as Macintosh serial ports) lack a true
 RTS output. Such serial ports use this mode to implement
 unidirectional flow control. The serial port will
 suspend transmission when requested by the modem (via CTS)
-but will be unable to request the modem stop sending to the
+but will be unable to request the modem to stop sending to the
 computer. This mode retains the ability to use DTR as
 a modem control line.
 .TP
@@ -100,8 +121,9 @@ This entry is removed when the PPP connection is broken.  This option
 is privileged if the \fInodefaultroute\fR option has been specified.
 .TP
 .B disconnect \fIscript
-Run the executable or shell command specified by \fIscript\fR after
-pppd has terminated the link.  This script could, for example, issue
+Execute the command specified by \fIscript\fR, by passing it to a
+shell, after
+pppd has terminated the link.  This command could, for example, issue
 commands to the modem to cause it to hang up if hardware modem control
 signals were not available.  The disconnect script is not run if the
 modem has already hung up.  A value for this option from a privileged
@@ -115,14 +137,14 @@ specified as a list of hex numbers separated by commas.  Note that
 almost any character can be specified for the \fIescape\fR option,
 unlike the \fIasyncmap\fR option which only allows control characters
 to be specified.  The characters which may not be escaped are those
-with hex values 0x20 - 0x3f or 0x5e.
+with hex values 0x20 \- 0x3f or 0x5e.
 .TP
 .B file \fIname
 Read options from file \fIname\fR (the format is described below).
 The file must be readable by the user who has invoked pppd.
 .TP
 .B init \fIscript
-Run the executable or shell command specified by \fIscript\fR to
+Execute the command specified by \fIscript\fR, by passing it to a shell, to
 initialize the serial line.  This script would typically use the
 chat(8) program to configure the modem to enable auto answer.  A value
 for this option from a privileged source cannot be overridden by a
@@ -134,17 +156,19 @@ serial device to ensure exclusive access to the device.
 .TP
 .B mru \fIn
 Set the MRU [Maximum Receive Unit] value to \fIn\fR. Pppd
-will ask the peer to send packets of no more than \fIn\fR bytes.  The
-minimum MRU value is 128.  The default MRU value is 1500.  A value of
-296 is recommended for slow links (40 bytes for TCP/IP header + 256
-bytes of data).  (Note that for IPv6 MRU must be at least 1280)
+will ask the peer to send packets of no more than \fIn\fR bytes.
+The value of \fIn\fR must be between 128 and 16384; the default is 1500.
+A value of
+296 works well on very slow links (40 bytes for TCP/IP header + 256
+bytes of data).
+Note that for the IPv6 protocol, the MRU must be at least 1280.
 .TP
 .B mtu \fIn
 Set the MTU [Maximum Transmit Unit] value to \fIn\fR.  Unless the
 peer requests a smaller value via MRU negotiation, pppd will
 request that the kernel networking code send data packets of no more
-than \fIn\fR bytes through the PPP network interface.  (Note that for 
-IPv6 MTU must be at least 1280)
+than \fIn\fR bytes through the PPP network interface.  Note that for 
+the IPv6 protocol, the MTU must be at least 1280.
 .TP
 .B passive
 Enables the "passive" option in the LCP.  With this option, pppd will
@@ -191,8 +215,11 @@ except that qualifiers which are inappropriate for a PPP link, such as
 \fBether\fR and \fBarp\fR, are not permitted.  Generally the filter
 expression should be enclosed in single-quotes to prevent whitespace
 in the expression from being interpreted by the shell. This option
-is currently only available under NetBSD, and then only
-if both the kernel and pppd were compiled with PPP_FILTER defined.
+is currently only available under Linux, and requires that the kernel
+was configured to include PPP filtering support (CONFIG_PPP_FILTER).
+Note that it
+is possible to apply different constraints to incoming and outgoing
+packets using the \fBinbound\fR and \fBoutbound\fR qualifiers.
 .TP
 .B allow-ip \fIaddress(es)
 Allow peers to use the given IP address or subnet without
@@ -200,6 +227,10 @@ authenticating themselves.  The parameter is parsed as for each
 element of the list of allowed IP addresses in the secrets files (see
 the AUTHENTICATION section below).
 .TP
+.B allow-number \fInumber
+Allow peers to connect from the given telephone number.  A trailing
+`*' character will match all numbers beginning with the leading part.
+.TP
 .B bsdcomp \fInr,nt
 Request that the peer compress packets that it sends, using the
 BSD-Compress scheme, with a maximum code size of \fInr\fR bits, and
@@ -235,8 +266,16 @@ Set the maximum number of CHAP challenge transmissions to \fIn\fR
 Set the CHAP restart interval (retransmission timeout for challenges)
 to \fIn\fR seconds (default 3).
 .TP
+.B child-timeout \fIn
+When exiting, wait for up to \fIn\fR seconds for any child processes
+(such as the command specified with the \fBpty\fR command) to exit
+before exiting.  At the end of the timeout, pppd will send a SIGTERM
+signal to any remaining child processes and exit.  A value of 0 means
+no timeout, that is, pppd will wait until all child processes have
+exited.
+.TP
 .B connect-delay \fIn
-Wait for up \fIn\fR milliseconds after the connect script finishes for
+Wait for up to \fIn\fR milliseconds after the connect script finishes for
 a valid PPP packet from the peer.  At the end of this time, or when a
 valid PPP packet is received from the peer, pppd will commence
 negotiation by sending its first LCP packet.  The default value is
@@ -327,6 +366,28 @@ the MAC type, the value may also be the name of an ethernet or similar
 network interface.  This option is currently only available under
 Linux.
 .TP
+.B eap-interval \fIn
+If this option is given and pppd authenticates the peer with EAP
+(i.e., is the server), pppd will restart EAP authentication every
+\fIn\fR seconds.  For EAP SRP-SHA1, see also the \fBsrp-interval\fR
+option, which enables lightweight rechallenge.
+.TP
+.B eap-max-rreq \fIn
+Set the maximum number of EAP Requests to which pppd will respond (as
+a client) without hearing EAP Success or Failure.  (Default is 20.)
+.TP
+.B eap-max-sreq \fIn
+Set the maximum number of EAP Requests that pppd will issue (as a
+server) while attempting authentication.  (Default is 10.)
+.TP
+.B eap-restart \fIn
+Set the retransmit timeout for EAP Requests when acting as a server
+(authenticator).  (Default is 3 seconds.)
+.TP
+.B eap-timeout \fIn
+Set the maximum time to wait for the peer to send an EAP Request when
+acting as a client (authenticatee).  (Default is 20 seconds.)
+.TP
 .B hide-password
 When logging the contents of PAP packets, this option causes pppd to
 exclude the password string from the log.  This is the default.
@@ -561,6 +622,10 @@ control, as for the \fIcrtscts\fR option.
 Enables the use of PPP multilink; this is an alias for the `multilink'
 option.  This option is currently only available under Linux.
 .TP
+.B mppe-stateful
+Allow MPPE to use stateful mode.  Stateless mode is still attempted first.
+The default is to disallow stateful mode.  
+.TP
 .B mpshortseq
 Enables the use of short (12-bit) sequence numbers in multilink
 headers, as opposed to 24-bit sequence numbers.  This option is only
@@ -607,17 +672,6 @@ will be used as the name to send to the peer when authenticating the
 local system to the peer.  (Note that pppd does not append the domain
 name to \fIname\fR.)
 .TP
-.B netmask \fIn
-Set the interface netmask to \fIn\fR, a 32 bit netmask in "decimal dot"
-notation (e.g. 255.255.255.0).  If this option is given, the value
-specified is ORed with the default netmask.  The default netmask is
-chosen based on the negotiated remote IP address; it is the
-appropriate network mask for the class of the remote IP address, ORed
-with the netmasks for any non point-to-point network interfaces in the
-system which are on the same network.  (Note: on some platforms, pppd
-will always use 255.255.255.255 for the netmask, if that is the only
-appropriate value for a point-to-point interface.)
-.TP
 .B noaccomp
 Disable Address/Control compression in both directions (send and
 receive).
@@ -703,6 +757,18 @@ peer is buggy.
 Disables the use of PPP multilink.  This option is currently only
 available under Linux.
 .TP
+.B nomppe
+Disables MPPE (Microsoft Point to Point Encryption).  This is the default.
+.TP
+.B nomppe-40
+Disable 40\-bit encryption with MPPE.
+.TP
+.B nomppe-128
+Disable 128\-bit encryption with MPPE.
+.TP
+.B nomppe-stateful
+Disable MPPE stateful mode.  This is the default.
+.TP
 .B nompshortseq
 Disables the use of short (12-bit) sequence numbers in the PPP
 multilink protocol, forcing the use of 24-bit sequence numbers.  This
@@ -776,7 +842,7 @@ Specifies a packet filter to applied to data packets being sent or
 received to determine which packets should be allowed to pass.
 Packets which are rejected by the filter are silently discarded.  This
 option can be used to prevent specific network daemons (such as
-routed) using up link bandwidth, or to provide a basic firewall
+routed) using up link bandwidth, or to provide a very basic firewall
 capability.
 The \fIfilter-expression\fR syntax is as described for tcpdump(1),
 except that qualifiers which are inappropriate for a PPP link, such as
@@ -785,8 +851,13 @@ expression should be enclosed in single-quotes to prevent whitespace
 in the expression from being interpreted by the shell.  Note that it
 is possible to apply different constraints to incoming and outgoing
 packets using the \fBinbound\fR and \fBoutbound\fR qualifiers. This
-option is currently only available under NetBSD, and then only if both
-the kernel and pppd were compiled with PPP_FILTER defined.
+option is currently only available under Linux, and requires that the
+kernel was configured to include PPP filtering support (CONFIG_PPP_FILTER).
+.TP
+.B password \fIpassword-string
+Specifies the password to use for authenticating to the peer.  Use
+of this option is discouraged, as the password is likely to be visible
+to other users on the system (for example, by using ps(1)).
 .TP
 .B persist
 Do not exit after a connection is terminated; instead try to reopen
@@ -849,6 +920,10 @@ displayed in readable form using the pppdump(8) program.
 Set the assumed name of the remote system for authentication purposes
 to \fIname\fR.
 .TP
+.B remotenumber \fInumber
+Set the assumed telephone number of the remote system for authentication
+purposes to \fInumber\fR.
+.TP
 .B refuse-chap
 With this option, pppd will not agree to authenticate itself to the
 peer using CHAP.
@@ -857,6 +932,14 @@ peer using CHAP.
 With this option, pppd will not agree to authenticate itself to the
 peer using MS-CHAP.
 .TP
+.B refuse-mschap-v2
+With this option, pppd will not agree to authenticate itself to the
+peer using MS-CHAPv2.
+.TP
+.B refuse-eap
+With this option, pppd will not agree to authenticate itself to the
+peer using EAP.
+.TP
 .B refuse-pap
 With this option, pppd will not agree to authenticate itself to the
 peer using PAP.
@@ -865,10 +948,32 @@ peer using PAP.
 Require the peer to authenticate itself using CHAP [Challenge
 Handshake Authentication Protocol] authentication.
 .TP
+.B require-mppe
+Require the use of MPPE (Microsoft Point to Point Encryption).  This
+option disables all other compression types.  This option enables
+both 40\-bit and 128\-bit encryption.  In order for MPPE to successfully
+come up, you must have authenticated with either MS-CHAP or MS-CHAPv2.
+This option is presently only supported under Linux, and only if your
+kernel has been configured to include MPPE support.
+.TP
+.B require-mppe-40
+Require the use of MPPE, with 40\-bit encryption.
+.TP
+.B require-mppe-128
+Require the use of MPPE, with 128\-bit encryption.
+.TP
 .B require-mschap
-Require the peer to authenticate itself using MS-CHAP [Microsft Challenge
+Require the peer to authenticate itself using MS-CHAP [Microsoft Challenge
 Handshake Authentication Protocol] authentication.
 .TP
+.B require-mschap-v2
+Require the peer to authenticate itself using MS-CHAPv2 [Microsoft Challenge
+Handshake Authentication Protocol, Version 2] authentication.
+.TP
+.B require-eap
+Require the peer to authenticate itself using EAP [Extensible
+Authentication Protocol] authentication.
+.TP
 .B require-pap
 Require the peer to authenticate itself using PAP [Password
 Authentication Protocol] authentication.
@@ -882,12 +987,37 @@ With this option, pppd will not transmit LCP packets to initiate a
 connection until a valid LCP packet is received from the peer (as for
 the `passive' option with ancient versions of pppd).
 .TP
+.B srp-interval \fIn
+If this parameter is given and pppd uses EAP SRP-SHA1 to authenticate
+the peer (i.e., is the server), then pppd will use the optional
+lightweight SRP rechallenge mechanism at intervals of \fIn\fR
+seconds.  This option is faster than \fBeap-interval\fR
+reauthentication because it uses a hash-based mechanism and does not
+derive a new session key.
+.TP
+.B srp-pn-secret \fIstring
+Set the long-term pseudonym-generating secret for the server.  This
+value is optional and if set, needs to be known at the server
+(authenticator) side only, and should be different for each server (or
+poll of identical servers).  It is used along with the current date to
+generate a key to encrypt and decrypt the client's identity contained
+in the pseudonym.
+.TP
+.B srp-use-pseudonym
+When operating as an EAP SRP-SHA1 client, attempt to use the pseudonym
+stored in ~/.ppp_psuedonym first as the identity, and save in this
+file any pseudonym offered by the peer during authentication.
+.TP
 .B sync
 Use synchronous HDLC serial encoding instead of asynchronous.
 The device used by pppd with this option must have sync support.
 Currently supports Microgate SyncLink adapters
 under Linux and FreeBSD 2.2.8 and later.
 .TP
+.B unit \fInum
+Sets the ppp unit number (for a ppp0 or ppp1 etc interface name) for outbound
+connections.
+.TP
 .B updetach
 With this option, pppd will detach from its controlling terminal once
 it has successfully established the ppp connection (to the point where
@@ -903,7 +1033,8 @@ the \fIname\fR option).  This option is not normally needed since the
 .B usepeerdns
 Ask the peer for up to 2 DNS server addresses.  The addresses supplied
 by the peer (if any) are passed to the /etc/ppp/ip-up script in the
-environment variables DNS1 and DNS2.  In addition, pppd will create an
+environment variables DNS1 and DNS2, and the environment variable
+USEPEERDNS will be set to 1.  In addition, pppd will create an
 /etc/ppp/resolv.conf file containing one or two nameserver lines with
 the address(es) supplied by the peer.
 .TP
@@ -998,15 +1129,18 @@ knows that secret.  Very often, the names used for authentication
 correspond to the internet hostnames of the peers, but this is not
 essential.
 .LP
-At present, pppd supports two authentication protocols: the Password
-Authentication Protocol (PAP) and the Challenge Handshake
-Authentication Protocol (CHAP).  PAP involves the client sending its
-name and a cleartext password to the server to authenticate itself.
-In contrast, the server initiates the CHAP authentication exchange by
-sending a challenge to the client (the challenge packet includes the
-server's name).  The client must respond with a response which
-includes its name plus a hash value derived from the shared secret and
-the challenge, in order to prove that it knows the secret.
+At present, pppd supports three authentication protocols: the Password
+Authentication Protocol (PAP), Challenge Handshake Authentication
+Protocol (CHAP), and Extensible Authentication Protocol (EAP).  PAP
+involves the client sending its name and a cleartext password to the
+server to authenticate itself.  In contrast, the server initiates the
+CHAP authentication exchange by sending a challenge to the client (the
+challenge packet includes the server's name).  The client must respond
+with a response which includes its name plus a hash value derived from
+the shared secret and the challenge, in order to prove that it knows
+the secret.  EAP supports CHAP-style authentication, and also includes
+the SRP-SHA1 mechanism, which is resistant to dictionary-based attacks
+and does not require a cleartext password on the server side.
 .LP
 The PPP protocol, being symmetrical, allows both peers to require the
 other to authenticate itself.  In that case, two separate and
@@ -1020,14 +1154,16 @@ pppd will not agree to authenticate itself with a particular protocol
 if it has no secrets which could be used to do so.
 .LP
 Pppd stores secrets for use in authentication in secrets
-files (/etc/ppp/pap-secrets for PAP, /etc/ppp/chap-secrets for CHAP/MS-CHAP).
-Both secrets files have the same format.  The secrets files can
+files (/etc/ppp/pap-secrets for PAP, /etc/ppp/chap-secrets for CHAP,
+MS-CHAP, MS-CHAPv2, and EAP MD5-Challenge, and /etc/ppp/srp-secrets
+for EAP SRP-SHA1).
+All secrets files have the same format.  The secrets files can
 contain secrets for pppd to use in authenticating itself to other
 systems, as well as secrets for pppd to use when authenticating other
 systems to itself.
 .LP
 Each line in a secrets file contains one secret.  A given secret is
-specific to a particular combination of client and server - it can
+specific to a particular combination of client and server \- it can
 only be used by that client to authenticate itself to that server.
 Thus each line in a secrets file has at least 3 fields: the name of
 the client, the name of the server, and the secret.  These fields may
@@ -1066,7 +1202,9 @@ field and the name of the local system in the second field.  The
 name of the local system defaults to the hostname, with the domain
 name appended if the \fIdomain\fR option is used.  This default can be
 overridden with the \fIname\fR option, except when the
-\fIusehostname\fR option is used.
+\fIusehostname\fR option is used.  (For EAP SRP-SHA1, see the
+srp-entry(8) utility for generating proper validator entries to be
+used in the "secret" field.)
 .LP
 When pppd is choosing a secret to use in authenticating itself to the
 peer, it first determines what name it is going to use to identify
@@ -1075,14 +1213,14 @@ itself to the peer.  This name can be specified by the user with the
 the name of the local system, determined as described in the previous
 paragraph.  Then pppd looks for a secret with this name in the first
 field and the peer's name in the second field.  Pppd will know the
-name of the peer if CHAP authentication is being used, because the
-peer will have sent it in the challenge packet.  However, if PAP is being
-used, pppd will have to determine the peer's name from the options
-specified by the user.  The user can specify the peer's name directly
-with the \fIremotename\fR option.  Otherwise, if the remote IP address
-was specified by a name (rather than in numeric form), that name will
-be used as the peer's name.  Failing that, pppd will use the null
-string as the peer's name.
+name of the peer if CHAP or EAP authentication is being used, because
+the peer will have sent it in the challenge packet.  However, if PAP
+is being used, pppd will have to determine the peer's name from the
+options specified by the user.  The user can specify the peer's name
+directly with the \fIremotename\fR option.  Otherwise, if the remote
+IP address was specified by a name (rather than in numeric form), that
+name will be used as the peer's name.  Failing that, pppd will use the
+null string as the peer's name.
 .LP
 When authenticating the peer with PAP, the supplied password is first
 compared with the secret from the secrets file.  If the password
@@ -1248,16 +1386,18 @@ pppd proxyarp
 .LP
 To allow a user to use the PPP facilities, you need to allocate an IP
 address for that user's machine and create an entry in
-/etc/ppp/pap-secrets or /etc/ppp/chap-secrets (depending on which
-authentication method the PPP implementation on the user's machine
-supports), so that the user's
-machine can authenticate itself.  For example, if Joe has a machine
-called "joespc" which is to be allowed to dial in to the machine
-called "server" and use the IP address joespc.my.net, you would add an
-entry like this to /etc/ppp/pap-secrets or /etc/ppp/chap-secrets:
+/etc/ppp/pap-secrets, /etc/ppp/chap-secrets, or /etc/ppp/srp-secrets
+(depending on which authentication method the PPP implementation on
+the user's machine supports), so that the user's machine can
+authenticate itself.  For example, if Joe has a machine called
+"joespc" that is to be allowed to dial in to the machine called
+"server" and use the IP address joespc.my.net, you would add an entry
+like this to /etc/ppp/pap-secrets or /etc/ppp/chap-secrets:
 .IP
 joespc server  "joe's secret"  joespc.my.net
 .LP
+(See srp-entry(8) for a means to generate the server's entry when
+SRP-SHA1 is in use.)
 Alternatively, you can create a username called (for example) "ppp",
 whose login shell is pppd and whose home directory is /etc/ppp.
 Options to be used when pppd is run this way can be put in
@@ -1276,13 +1416,15 @@ stream.
 .SH DIAGNOSTICS
 .LP
 Messages are sent to the syslog daemon using facility LOG_DAEMON.
-(This can be overriden by recompiling pppd with the macro
-LOG_PPP defined as the desired facility.)  In order to see the error
-and debug messages, you will need to edit your /etc/syslog.conf file
-to direct the messages to the desired output device or file.
+(This can be overridden by recompiling pppd with the macro
+LOG_PPP defined as the desired facility.)  See the syslog(8)
+documentation for details of where the syslog daemon will write the
+messages.  On most systems, the syslog daemon uses the
+/etc/syslog.conf file to specify the destination(s) for syslog
+messages.  You may need to edit that file to suit.
 .LP
 The \fIdebug\fR option causes the contents of all control packets sent
-or received to be logged, that is, all LCP, PAP, CHAP or IPCP packets.
+or received to be logged, that is, all LCP, PAP, CHAP, EAP, or IPCP packets.
 This can be useful if the PPP negotiation does not succeed or if
 authentication fails.
 If debugging is enabled at compile time, the \fIdebug\fR option also
@@ -1422,6 +1564,14 @@ the connection.
 .TP
 .B LINKNAME
 The logical name of the link, set with the \fIlinkname\fR option.
+.TP
+.B DNS1
+If the peer supplies DNS server addresses, this variable is set to the
+first DNS server address supplied.
+.TP
+.B DNS2
+If the peer supplies DNS server addresses, this variable is set to the
+second DNS server address supplied.
 .P
 Pppd invokes the following scripts, if they exist.  It is not an error
 if they don't exist.
@@ -1498,7 +1648,8 @@ script.
 .B /var/run/ppp\fIn\fB.pid \fR(BSD or Linux), \fB/etc/ppp/ppp\fIn\fB.pid \fR(others)
 Process-ID for pppd process on ppp interface unit \fIn\fR.
 .TP
-.B /var/run/ppp-\fIname\fB.pid \fR(BSD or Linux), \fB/etc/ppp/ppp-\fIname\fB.pid \fR(others)
+.B /var/run/ppp-\fIname\fB.pid \fR(BSD or Linux),
+\fB/etc/ppp/ppp-\fIname\fB.pid \fR(others)
 Process-ID for pppd process for logical link \fIname\fR (see the
 \fIlinkname\fR option).
 .TP
@@ -1508,11 +1659,21 @@ file should be owned by root and not readable or writable by any other
 user.  Pppd will log a warning if this is not the case.
 .TP
 .B /etc/ppp/chap-secrets
-Names, secrets and IP addresses for CHAP/MS-CHAP authentication.  As for
+Names, secrets and IP addresses for CHAP/MS-CHAP/MS-CHAPv2 authentication.
+As for /etc/ppp/pap-secrets, this file should be owned by root and not
+readable or writable by any other user.  Pppd will log a warning if
+this is not the case.
+.TP
+.B /etc/ppp/srp-secrets
+Names, secrets, and IP addresses for EAP authentication.  As for
 /etc/ppp/pap-secrets, this file should be owned by root and not
 readable or writable by any other user.  Pppd will log a warning if
 this is not the case.
 .TP
+.B ~/.ppp_pseudonym
+Saved client-side SRP-SHA1 pseudonym.  See the \fIsrp-use-pseudonym\fR
+option for details.
+.TP
 .B /etc/ppp/options
 System default options for pppd, read before user default options or
 command-line options.
@@ -1565,12 +1726,28 @@ Simpson, W.A.
 .I PPP in HDLC-like Framing.
 July 1994.
 .TP
+.B RFC2284
+Blunk, L.; Vollbrecht, J.,
+.I PPP Extensible Authentication Protocol (EAP).
+March 1998.
+.TP
 .B RFC2472
 Haskin, D.
 .I IP Version 6 over PPP
 December 1998.
+.TP
+.B RFC2945
+Wu, T.,
+.I The SRP Authentication and Key Exchange System
+September 2000.
+.TP
+.B draft-ietf-pppext-eap-srp-03.txt
+Carlson, J.; et al.,
+.I EAP SRP-SHA1 Authentication Protocol.
+July 2001.
 .SH NOTES
-The following signals have the specified effect when sent to pppd.
+Some limited degree of control can be exercised over a running pppd
+process by sending it a signal from the list below.
 .TP
 .B SIGINT, SIGTERM
 These signals cause pppd to terminate the link (by closing LCP),
@@ -1594,10 +1771,101 @@ of a fatal decompression error.  (Fatal decompression errors generally
 indicate a bug in one or other implementation.)
 
 .SH AUTHORS
-Paul Mackerras (Paul.Mackerras@cs.anu.edu.au), based on earlier work by
+Paul Mackerras (Paul.Mackerras@samba.org), based on earlier work by
 Drew Perkins,
 Brad Clements,
 Karl Fox,
 Greg Christy,
 and
 Brad Parker.
+
+.SH COPYRIGHT
+Pppd is copyrighted and made available under conditions which provide
+that it may be copied and used in source or binary forms provided that
+the conditions listed below are met.  Portions of pppd are covered by
+the following copyright notices:
+.LP
+Copyright (c) 1984-2000 Carnegie Mellon University. All rights
+reserved.
+.br
+Copyright (c) 1993-2004 Paul Mackerras. All rights reserved.
+.br
+Copyright (c) 1995 Pedro Roque Marques.  All rights reserved.
+.br
+Copyright (c) 1995 Eric Rosenquist.  All rights reserved.
+.br
+Copyright (c) 1999 Tommi Komulainen.  All rights reserved.
+.br
+Copyright (C) Andrew Tridgell 1999
+.br
+Copyright (c) 2000 by Sun Microsystems, Inc.  All rights reserved.
+.br
+Copyright (c) 2001 by Sun Microsystems, Inc.  All rights reserved.
+.br
+Copyright (c) 2002 Google, Inc.  All rights reserved.
+.LP
+The copyright notices contain the following statements.
+.LP
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions
+are met:
+.LP
+1. Redistributions of source code must retain the above copyright
+   notice, this list of conditions and the following disclaimer.
+.LP
+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.
+.LP
+3. The name "Carnegie Mellon University" must not be used to
+   endorse or promote products derived from this software without
+   prior written permission. For permission or any legal
+   details, please contact
+.br
+     Office of Technology Transfer
+.br
+     Carnegie Mellon University
+.br
+     5000 Forbes Avenue
+.br
+     Pittsburgh, PA  15213-3890
+.br
+     (412) 268-4387, fax: (412) 268-7395
+.br
+     tech-transfer@andrew.cmu.edu
+.LP
+3b. The name(s) of the authors of this software must not be used to
+   endorse or promote products derived from this software without
+   prior written permission.
+.LP
+4. Redistributions of any form whatsoever must retain the following
+   acknowledgments:
+.br
+   "This product includes software developed by Computing Services
+    at Carnegie Mellon University (http://www.cmu.edu/computing/)."
+.br
+   "This product includes software developed by Paul Mackerras
+    <paulus@samba.org>".
+.br
+   "This product includes software developed by Pedro Roque Marques
+    <pedro_m@yahoo.com>".
+.br
+   "This product includes software developed by Tommi Komulainen
+    <Tommi.Komulainen@iki.fi>".
+.LP
+CARNEGIE MELLON UNIVERSITY DISCLAIMS ALL WARRANTIES WITH REGARD TO
+THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY
+AND FITNESS, IN NO EVENT SHALL CARNEGIE MELLON UNIVERSITY BE LIABLE
+FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN
+AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING
+OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+.LP
+THE AUTHORS OF THIS SOFTWARE DISCLAIM ALL WARRANTIES WITH REGARD TO
+THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY
+AND FITNESS, IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY
+SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN
+AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING
+OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.