GPRS / pppd

Hello World,

I've got a few problems / questions about dialing into a gprs network.


Here are my files:

1. /etc/ppp/options.gprs:
/dev/modem
debug
115200
defaultroute
ipcp-accept-local
ipcp-accept-remote
noauth
novj
connect 'chat -v -f chat.gprs'
user (E-Mail Removed)
usepeerdns
mru 1500
mtu 1500

2. /etc/ppp/chat.gprs:
TIMEOUT 5
ECHO ON
ABORT '\nBUSY\r'
ABORT '\nERROR\r'
ABORT '\nNO ANSWER\r'
ABORT '\nNO CARRIER\r'
ABORT '\nNO DIALTONE\r'
ABORT '\nRINGING\r\n\r\nRINGING\r'
'' \rAT
TIMEOUT 12
OK ATE1
OK 'AT+CPIN=3509'
OK 'AT+cgdcont=1,"IP","A1.net"'
OK ATD*99***1#
CONNECT '\d\c'

3. /etc/ppp/chap-secrets:
"(E-Mail Removed)" * "test"


When trying to dial-in with the scripts above the peer rejects me with
the follwing error:

linux:/etc/ppp # pppd file /etc/ppp/options.gprs
AT
OK
ATE1
OK
AT+cgdcont=1,"IP","A1.net"
OK
Serial connection established.
using channel 21
Using interface ppp0
Connect: ppp0 <--> /dev/modem
rcvd [LCP ConfReq id=0x1 <mru 1500> <asyncmap 0x0> <pcomp> <accomp>
<auth pap>]
sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0xea8db58a> <pcomp>
<accomp>]
sent [LCP ConfNak id=0x1 <auth eap>]
rcvd [LCP ConfRej id=0x1 <magic 0xea8db58a>]
sent [LCP ConfReq id=0x2 <asyncmap 0x0> <pcomp> <accomp>]
rcvd [LCP ConfReq id=0x2 <mru 1500> <asyncmap 0x0> <pcomp> <accomp>
<auth pap>]
sent [LCP ConfNak id=0x2 <auth eap>]
rcvd [LCP ConfAck id=0x2 <asyncmap 0x0> <pcomp> <accomp>]
rcvd [LCP ConfReq id=0x3 <mru 1500> <asyncmap 0x0> <pcomp> <accomp>
<auth pap>]
sent [LCP ConfNak id=0x3 <auth eap>]
rcvd [LCP ConfReq id=0x4 <mru 1500> <asyncmap 0x0> <pcomp> <accomp>
<auth pap>]
sent [LCP ConfNak id=0x4 <auth eap>]
rcvd [LCP ConfReq id=0x5 <mru 1500> <asyncmap 0x0> <pcomp> <accomp>
<auth pap>]
sent [LCP ConfNak id=0x5 <auth eap>]
rcvd [LCP ConfReq id=0x6 <mru 1500> <asyncmap 0x0> <pcomp> <accomp>
<auth pap>]
sent [LCP ConfRej id=0x6 <auth pap>]
rcvd [LCP ConfReq id=0x7 <mru 1500> <asyncmap 0x0> <pcomp> <accomp>]
sent [LCP ConfAck id=0x7 <mru 1500> <asyncmap 0x0> <pcomp> <accomp>]
sent [LCP EchoReq id=0x0 magic=0x0]
sent [IPCP ConfReq id=0x1 <addr 0.0.0.0> <ms-dns1 0.0.0.0> <ms-dns3
0.0.0.0>]
rcvd [LCP TermReq id=0x8]
LCP terminated by peer
sent [LCP TermAck id=0x8]
rcvd [LCP EchoRep id=0x0 magic=0x0]
Modem hangup
Connection terminated.



I am not sure what that really means and I am irritated, because this
rejection only happens with one of my two modems:
It happens with my WaveCom Wismo Quik Q2501B (on baudrate 115200) but
not with my Falcom Twist (on baudrate 19200). I used exactly the same
scripts (except for the baudrate in "options.gprs").


Another question is:
How to submit a pin-code within the script?

If I re-submit the pin-code, the script fails:

linux:/etc/ppp # pppd file /etc/ppp/options.gprs
AT
OK
ATE1
OK
AT+CPIN=3509
ERROR
Connect script failed


I know that "at+cpin?" tells me if the pincode was alread submitted,
but I don't know how to build that into my scripts.

I would appreciate if someone was able to help me

yours

Herwig

On 17 Jul 2005 10:25:09 -0700, "132.a.dabei" <(E-Mail Removed)>
wrote:


>I am not sure what that really means and I am irritated, because this
>rejection only happens with one of my two modems:

It's probably the modem, or sim card. Do you use the same sim card?


>It happens with my WaveCom Wismo Quik Q2501B (on baudrate 115200) but
>not with my Falcom Twist (on baudrate 19200). I used exactly the same
>scripts (except for the baudrate in "options.gprs").

I was never very happy with Wavecom, although it's been a long time
since I used one. I would have hope they had got over their problems.


>Another question is:
>How to submit a pin-code within the script?
>
>If I re-submit the pin-code, the script fails:
>
>linux:/etc/ppp # pppd file /etc/ppp/options.gprs
>AT
>OK
>ATE1
>OK

>AT+CPIN=3509

I think you need quotation marks around the pin number:
at+cpin="3509"


>I know that "at+cpin?" tells me if the pincode was alread submitted,
>but I don't know how to build that into my scripts.

I had trouble getting a script to work for me. I ended up writing a
C++ program to do the modem initialisation.

Dan

Hi dan,

thank you for your suggestions.
It wasn't the sim-card. I had to rename the file "chap-secrets" into
"pap-secrets". That's kinda strange, because I really used the same sim
!?!?!

However: Using quotation marks around the pin code makes no change: the
script fails.

C++: I also started to write a program in c++: Sending AT commands to
/dev/ttyS0 is something easy. The point where I got stuck was: I had no
idea how to start the ppp connection after I initialized the modem. How
did you do that? If I knew how to implement ppp into my project, it
would be the best solution.

Thank you

Herwig

On 18 Jul 2005 00:08:35 -0700, "132.a.dabei" <(E-Mail Removed)>
wrote:

>The point where I got stuck was: I had no
>idea how to start the ppp connection after I initialized the modem. How
>did you do that? If I knew how to implement ppp into my project, it
>would be the best solution.


I'm trying to remember what I did. :-)

In the /etc/ppp/peers directory I have a ppp options file called
'gprs'. In that file, one of the options is:
connect /bin/g18
which is the C++ modem startup program. So pppd runs the modem
startup program itself.

Pppd runs the ip-up script after it's connected. I copy the
/etc/ppp/resolv.conf that pppd updated into /etc/resolv.conf

Have a good read of man pppd.

Dan

132.a.dabei <(E-Mail Removed)> wrote:
> Hello World,

> I've got a few problems / questions about dialing into a gprs network.


> Here are my files:

> 1. /etc/ppp/options.gprs:
> /dev/modem
> debug
> 115200
> defaultroute
> ipcp-accept-local
> ipcp-accept-remote
> noauth
> novj
> connect 'chat -v -f chat.gprs'
> user (E-Mail Removed)
> usepeerdns
> mru 1500
> mtu 1500

> 2. /etc/ppp/chat.gprs:
> TIMEOUT 5
> ECHO ON
> ABORT '\nBUSY\r'
> ABORT '\nERROR\r'
> ABORT '\nNO ANSWER\r'
> ABORT '\nNO CARRIER\r'
> ABORT '\nNO DIALTONE\r'
> ABORT '\nRINGING\r\n\r\nRINGING\r'
> '' \rAT
> TIMEOUT 12
> OK ATE1
> OK 'AT+CPIN=3509'
> OK 'AT+cgdcont=1,"IP","A1.net"'
> OK ATD*99***1#
> CONNECT '\d\c'

> 3. /etc/ppp/chap-secrets:
> "(E-Mail Removed)" * "test"


> When trying to dial-in with the scripts above the peer rejects me with
> the follwing error:

> linux:/etc/ppp # pppd file /etc/ppp/options.gprs
> AT
> OK
> ATE1
> OK
> AT+cgdcont=1,"IP","A1.net"
> OK
> Serial connection established.
> using channel 21
> Using interface ppp0
> Connect: ppp0 <--> /dev/modem
> rcvd [LCP ConfReq id=0x1 <mru 1500> <asyncmap 0x0> <pcomp> <accomp>
> <auth pap>]
> sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0xea8db58a> <pcomp>
> <accomp>]
> sent [LCP ConfNak id=0x1 <auth eap>]

The peer wants to authenticate you using PAP, but pppd wants it to use
EAP instead. The peer in this case may be a PPP implementation inside
the "modem" (whether the PPP negotiations are sent elsewhere over GPRS
for processing I don't know). The Nak suggesting EAP likely implies
/etc/ppp/srp-secrets exists and that the pppd refuse-pap and refuse-chap
options are used or there exist no {chap,pap}-secrets files. "Likely"
because EAP (Extensible Authentication Prococol) is new and I have no
experience with it.

In fact pppd _insists_ on EAP and eventually that leads to it terminating
the link.

Why you succeed using a different "modem" with the same host and pppd
options, except for speed, I cannot say - unless the successful one
provides for EAP in it's PPP implementation and the other does not.
(And I know nothing about GPRS pincodes.)

--
Clifford Kite Email: "echo xvgr_yvahk-(E-Mail Removed)|rot13"

132.a.dabei <(E-Mail Removed)> wrote:
> Hi dan,

> thank you for your suggestions.
> It wasn't the sim-card. I had to rename the file "chap-secrets" into
> "pap-secrets". That's kinda strange, because I really used the same sim
> !?!?!

It's not strange if the previously successful modem has an EAP
implementation that uses CHAP.

> However: Using quotation marks around the pin code makes no change: the
> script fails.

You might try adding "OK ATZ" just after "TIMEOUT 12" in the chat
script. That should flush modem buffers and reset the default profile.

> C++: I also started to write a program in c++: Sending AT commands to
> /dev/ttyS0 is something easy. The point where I got stuck was: I had no
> idea how to start the ppp connection after I initialized the modem. How
> did you do that? If I knew how to implement ppp into my project, it
> would be the best solution.

A wrapper containing the AT code followed by an execle command to start
"<directory-spec>pppd file /etc/ppp/options.gprs" should work in C -
dunno about C++.

--
Clifford Kite Email: "echo xvgr_yvahk-(E-Mail Removed)|rot13"
PPP-Q&A links, downloads: http://ckite.no-ip.net/
/* "Be liberal in what you accept, and conservative in what you send"
RFC 1122 */

Clifford Kite <(E-Mail Removed)> writes:


>You might try adding "OK ATZ" just after "TIMEOUT 12" in the chat
>script. That should flush modem buffers and reset the default profile.

Uh, use AT&F (or AT&F1 if it is a Sportster modem). This loads the factory
defaults -- ATZ is a crap shoot. It
depends entirely on what setup was saved by some person or some program
sometime in the past.

Unruh <unruh-(E-Mail Removed)> wrote:
> Clifford Kite <(E-Mail Removed)> writes:

>>You might try adding "OK ATZ" just after "TIMEOUT 12" in the chat
>>script. That should flush modem buffers and reset the default profile.

> Uh, use AT&F (or AT&F1 if it is a Sportster modem). This loads the factory
> defaults -- ATZ is a crap shoot. It
> depends entirely on what setup was saved by some person or some program
> sometime in the past.

The idea behind the suggestion to use ATZ was to flush the "modem"
buffers where something left-over from previous usage might cause
the ERROR. An old modem manual claims ATZ does this and resets
the modem to it's defaults. His chat script didn't show any AT&F
initialization so I didn't suggest that "OK AT&F" also be added.

Of course, this is GPRS "modem" and the old Hayes de-facto standard
can only serve as a guide when you don't have a modem command manual.
So ATZ may not work and/or also result in an ERROR.

--
Clifford Kite Email: "echo xvgr_yvahk-(E-Mail Removed)|rot13"
PPP-Q&A links, downloads: http://ckite.no-ip.net/
/* Better is the enemy of good enough. */

Clifford Kite <(E-Mail Removed)> writes:

>Unruh <unruh-(E-Mail Removed)> wrote:
>> Clifford Kite <(E-Mail Removed)> writes:

>>>You might try adding "OK ATZ" just after "TIMEOUT 12" in the chat
>>>script. That should flush modem buffers and reset the default profile.

>> Uh, use AT&F (or AT&F1 if it is a Sportster modem). This loads the factory
>> defaults -- ATZ is a crap shoot. It
>> depends entirely on what setup was saved by some person or some program
>> sometime in the past.

>The idea behind the suggestion to use ATZ was to flush the "modem"
>buffers where something left-over from previous usage might cause
>the ERROR. An old modem manual claims ATZ does this and resets
>the modem to it's defaults. His chat script didn't show any AT&F
>initialization so I didn't suggest that "OK AT&F" also be added.

>Of course, this is GPRS "modem" and the old Hayes de-facto standard
>can only serve as a guide when you don't have a modem command manual.
>So ATZ may not work and/or also result in an ERROR.

Fine, except as we both know, ATZ often is implimented to recall a prior
saved setup rather than factory defaults, and can itself be the source of
grief when those saved setting are insane (which often happens). Usually
(Hayes ) AT&F sets up factor default (except on the sportster modems where
AT&F sets up software flow control, and you have to use AT&F1) and is
usually safer. However, as you point out, the modem manufacturer can do
what they damn well please and they may have AT&F igniting a charge inside
the modem. (LG did that with their CD drives. They used a command that the
standards said should be invalid on a Read only drive-- reserved for RW
drives-- and instead had it destroy the PROM memory on the read only drive. )

Ie, I would advise him to use AT&F to start with. If that is an invalid
command, use ATZ. Of course reading the manual, if it exists is an even
better idea.

Unruh <unruh-(E-Mail Removed)> wrote:
> Clifford Kite <(E-Mail Removed)> writes:
>>Unruh <unruh-(E-Mail Removed)> wrote:

>>> Uh, use AT&F (or AT&F1 if it is a Sportster modem). This loads the factory
>>> defaults -- ATZ is a crap shoot. It
>>> depends entirely on what setup was saved by some person or some program
>>> sometime in the past.

>>The idea behind the suggestion to use ATZ was to flush the "modem"
>>buffers where something left-over from previous usage might cause
>>the ERROR. An old modem manual claims ATZ does this and resets
>>the modem to it's defaults. His chat script didn't show any AT&F
>>initialization so I didn't suggest that "OK AT&F" also be added.

>>Of course, this is GPRS "modem" and the old Hayes de-facto standard
>>can only serve as a guide when you don't have a modem command manual.
>>So ATZ may not work and/or also result in an ERROR.

> Fine, except as we both know, ATZ often is implimented to recall a prior
> saved setup rather than factory defaults, and can itself be the source of
> grief when those saved setting are insane (which often happens).

Point taken, and it certainly won't hurt to try AT&F and/or AT&F1
after ATZ, if that alone doesn't work.

As you said, the Sportster requires AT&F1 for RTS/CTS flow-control
- but it's the only (normal) modem I've ever known about that used
XON/XOFF in the factory default profile. I'm also not as certain as
you are that a profile restored by ATZ (or ATZ[0,1]) is often a bad
user-created profile, but readily admit I can't say it isn't either.

--
Clifford Kite Email: "echo xvgr_yvahk-(E-Mail Removed)|rot13"