dial-in server problem

Please help I can't run my dial-in server on Suse 8.2.
I keep getting "LCP terminated by peer" and the ip-up script doesn't start
at all.
I'm dialing from WindowsNT and here is the log from /var/log/messages

Jan 19 10:01:15 linux pppd[8399]: pppd 2.4.1 started by LOGIN, uid 0
Jan 19 10:01:15 linux pppd[8399]: Perms of /dev/ttyS1 are ok, no 'mesg n'
neccesary.
Jan 19 10:01:15 linux pppd[8399]: using channel 7
Jan 19 10:01:15 linux pppd[8399]: Using interface ppp0
Jan 19 10:01:15 linux pppd[8399]: Connect: ppp0 <--> /dev/ttyS1
Jan 19 10:01:15 linux pppd[8399]: sent [LCP ConfReq id=0x1 <asyncmap 0x0>
<auth pap> <magic 0x8a68104d> <pcomp> <accomp>]
Jan 19 10:01:15 linux pppd[8399]: rcvd [LCP ConfRej id=0x1 <pcomp> <accomp>]
Jan 19 10:01:15 linux pppd[8399]: sent [LCP ConfReq id=0x2 <asyncmap 0x0>
<auth pap> <magic 0x8a68104d>]
Jan 19 10:01:16 linux pppd[8399]: rcvd [LCP ConfAck id=0x2 <asyncmap 0x0>
<auth pap> <magic 0x8a68104d>]
Jan 19 10:01:17 linux pppd[8399]: sent [LCP ConfReq id=0x2 <asyncmap 0x0>
<auth pap> <magic 0x8a68104d>]
Jan 19 10:01:18 linux pppd[8399]: rcvd [LCP ConfAck id=0x2 <asyncmap 0x0>
<auth pap> <magic 0x8a68104d>]
Jan 19 10:01:18 linux pppd[8399]: rcvd [LCP ConfReq id=0x2 <magic 0x29ed>
<callback CBCP> <asyncmap 0x0>]
Jan 19 10:01:18 linux pppd[8399]: sent [LCP ConfAck id=0x2 <magic 0x29ed>
<callback CBCP> <asyncmap 0x0>]
Jan 19 10:01:18 linux pppd[8399]: sent [LCP EchoReq id=0x0 magic=0x8a68104d]
Jan 19 10:01:18 linux pppd[8399]: cbcp_lowerup
Jan 19 10:01:18 linux pppd[8399]: want: 2
Jan 19 10:01:18 linux pppd[8399]: rcvd [LCP EchoRep id=0x0 magic=0x29ed]
Jan 19 10:01:19 linux pppd[8399]: rcvd [PAP AuthReq id=0x3 user="borut"
password=<hidden>]
Jan 19 10:01:19 linux pppd[8399]: user borut logged in
Jan 19 10:01:19 linux pppd[8399]: Script /etc/ppp/auth-up finished (pid
8432), status = 0x0
Jan 19 10:01:36 linux pppd[8399]: LCP terminated by peer
Jan 19 10:01:36 linux pppd[8399]: Script /etc/ppp/auth-down finished (pid
8435), status = 0x0
Jan 19 10:01:37 linux pppd[8399]: Hangup (SIGHUP)
Jan 19 10:01:37 linux pppd[8399]: Modem hangup
Jan 19 10:01:37 linux pppd[8399]: Connection terminated.
Jan 19 10:01:37 linux pppd[8399]: Connect time 0.4 minutes.
Jan 19 10:01:37 linux pppd[8399]: Sent 0 bytes, received 0 bytes.
Jan 19 10:01:37 linux pppd[8399]: Exit.


I have this settings in my configuration files\:

1)/etc/inittab

mo:235:respawn:/usr/sbin/mgetty -x 9 -D -s 57600 ttyS1

2)/etc/mgetty+sendfax/login.config

/AutoPPP/ - a_ppp /usr/sbin/pppd file /etc/ppp/options.server

3)/etc/mgetty+sendfax/mgetty.config
port-owner uucp
port-group uucp
port-mode 0664

4)/etc/ppp/options.server

auth
+pap
-chap
login
lock
modem
crtscts
debug
proxyarp
asyncmap 0
defaultroute
replacedefaultroute
noipdefault
noipx
kdebug 4
ms-dns 10.1.0.200
ms-wins 10.1.0.200

and the default settings:
nodetach
lcp-echo-interval 30
lcp-echo-failure 4
lcp-max-configure 60
lcp-restart 2
idle 600
5)/etc/ppp/pap-secrets

* * ""

Borut <(E-Mail Removed)> wrote:
> Please help I can't run my dial-in server on Suse 8.2.
> I keep getting "LCP terminated by peer" and the ip-up script doesn't start
> at all.
> I'm dialing from WindowsNT and here is the log from /var/log/messages

> Jan 19 10:01:15 linux pppd[8399]: pppd 2.4.1 started by LOGIN, uid 0
> Jan 19 10:01:15 linux pppd[8399]: Perms of /dev/ttyS1 are ok, no 'mesg n'
> neccesary.

This pppd is not the standard version. The message above doesn't exist
for the standard pppd.

> Jan 19 10:01:15 linux pppd[8399]: using channel 7
> Jan 19 10:01:15 linux pppd[8399]: Using interface ppp0
> Jan 19 10:01:15 linux pppd[8399]: Connect: ppp0 <--> /dev/ttyS1
> Jan 19 10:01:15 linux pppd[8399]: sent [LCP ConfReq id=0x1 <asyncmap 0x0>
> <auth pap> <magic 0x8a68104d> <pcomp> <accomp>]
> Jan 19 10:01:15 linux pppd[8399]: rcvd [LCP ConfRej id=0x1 <pcomp> <accomp>]
> Jan 19 10:01:15 linux pppd[8399]: sent [LCP ConfReq id=0x2 <asyncmap 0x0>
> <auth pap> <magic 0x8a68104d>]
> Jan 19 10:01:16 linux pppd[8399]: rcvd [LCP ConfAck id=0x2 <asyncmap 0x0>
> <auth pap> <magic 0x8a68104d>]
> Jan 19 10:01:17 linux pppd[8399]: sent [LCP ConfReq id=0x2 <asyncmap 0x0>
> <auth pap> <magic 0x8a68104d>]
> Jan 19 10:01:18 linux pppd[8399]: rcvd [LCP ConfAck id=0x2 <asyncmap 0x0>
> <auth pap> <magic 0x8a68104d>]
> Jan 19 10:01:18 linux pppd[8399]: rcvd [LCP ConfReq id=0x2 <magic 0x29ed>
> <callback CBCP> <asyncmap 0x0>]
> Jan 19 10:01:18 linux pppd[8399]: sent [LCP ConfAck id=0x2 <magic 0x29ed>
> <callback CBCP> <asyncmap 0x0>]

This pppd accepts a CBCP request, so the NT expects pppd to call it
back after the NT authenticates itself and hangs up. A standard pppd
only implements the "client" side of CBCP, not the server side and would
reject such a request. If you do not expect to do this then get and use
a standard pppd rather this one. The standard pppd source is available
at ftp.samba.org in /etc/ppp. (I'd recommend getting pppd 2.4.1 rather
than the newer 2.4.2 since I've had a problem with 2.4.2 that appears
to be caused by a bug in 2.4.2.)

If you do expect to call the NT back then you'll likely need a script
that answers the NT call and that also calls the NT back. Since a
number isn't included in the NT CBCP request, you will have to already
know the number to call.

The fact that pppd accepts the CBCP request reminds me of Red Hat 7.2
where RH tried to implement the server side in it's version of pppd
and broke it. It smells to me like the same thing is happening here.

-- Clifford Kite Email: "echo xvgr_yvahk-(E-Mail Removed)|rot13"
PPP-Q&A links, downloads: http://ckite.no-ip.net/
/* All the world's a net! And all the data in it merely packets
come to store-and-forward in the queues a while and then are
heard no more. 'Tis a network waiting to be switched!
-Vint Cerf, RFC 1121 */

Clifford Kite <(E-Mail Removed)> wrote:
> The standard pppd source is available
> at ftp.samba.org in /etc/ppp. (I'd recommend getting pppd 2.4.1 rather
> than the newer 2.4.2 since I've had a problem with 2.4.2 that appears
> to be caused by a bug in 2.4.2.)

For whomever might be interested, it was a bug. Paul Mackerras replied to
my query on the Linux PPP mailing list with a simple patch that fixed it.
The fix is to move a kill(0, sig); up one line in main.c:

~/shop/pppd/ppp-2.4.2$ diff -u -7 pppd/{main.c,main.c-original}
--- pppd/main.c Sat Jan 24 19:49:16 2004
+++ pppd/main.c-original Mon Jan 12 22:00:34 2004
@@ -1334,16 +1334,16 @@
kill_my_pg(sig)
int sig;
{
struct sigaction act, oldact;

act.sa_handler = SIG_IGN;
act.sa_flags = 0;
- kill(0, sig);
sigaction(sig, &act, &oldact);
+ kill(0, sig);
sigaction(sig, &oldact, NULL);
}

--
Clifford Kite Email: "echo xvgr_yvahk-(E-Mail Removed)|rot13"
PPP-Q&A links, downloads: http://ckite.no-ip.net/