[SR-Users] Kamailio - Nonce validity between Register and Invite

[Daniel-Constantin Mierla]

Hello,


On 8/31/12 5:08 PM, Daniel-Constantin Mierla wrote:
> Hello,
>
> I will look over the code, I just wanted to point that it is safer to 
> print by size in this case, otherwise it can get coredump, even if the 
> printed characters are not ascii, like:
>
> DBG("auth:pre_auth:api.c:nonce.c Invalid nonce (nonce %.*s)\n", 
> &b_nonce.n.md5_1[0], 16);

actually the size has to be first, like:

DBG("auth:pre_auth:api.c:nonce.c Invalid nonce (nonce %.*s)\n", 16, 
&b_nonce.n.md5_1[0]);

Cheers,
Daniel

>
>
> Otherwise DBG(..) goes on until finds '\0', which can be in a memory 
> zone used by another process.
>
> Cheers,
> Daniel
>
> On 8/31/12 4:09 PM, patrice.bodeven at orange.com wrote:
>> Hello
>>
>> The return code is 2.
>>
>> The problem seems due to "nonce.c"
>> if (!memcmp(&b_nonce.n.md5_1[0], &b_nonce2.n.md5_1[0], 16)) {
>> xxx
>> }
>>
>> So I put
>>
>>     DBG("auth:pre_auth:api.c:nonce.c Invalid nonce (nonce %s)\n", 
>> &b_nonce.n.md5_1[0]);
>>     DBG("auth:pre_auth:api.c:nonce.c Invalid nonce (nonce2 %s)\n", 
>> &b_nonce2.n.md5_1[0]);
>>     return 2;
>> }
>>
>> The result into log is :
>>
>> Aug 31 13:58:25 openims-kamailio-1 /usr/sbin/kamailio[7824]: DEBUG: 
>> registrar [lookup.c:229]: '+33296053686' found in usrloc
>> Aug 31 13:58:25 openims-kamailio-1 /usr/sbin/kamailio[7824]: DEBUG: 
>> auth_db [authorize.c:271]: realm value [openims-kamailio-1.mycluster]
>> Aug 31 13:58:25 openims-kamailio-1 /usr/sbin/kamailio[7824]: DEBUG: 
>> auth [api.c:95]: auth: digest-algo: MD5 parsed value: 1
>> Aug 31 13:58:25 openims-kamailio-1 /usr/sbin/kamailio[7824]: DEBUG: 
>> auth [nonce.c:473]: auth:pre_auth:api.c:nonce.c Invalid nonce (nonce 
>> ö§£Â\ÆX`#035~MT~\ÞUl~X~Sb~@»êü#025nÿ| ¹»©Zo#032#YåP~@Ãô#010)
>> Aug 31 13:58:25 openims-kamailio-1 /usr/sbin/kamailio[7824]: DEBUG: 
>> auth [nonce.c:474]: auth:pre_auth:api.c:nonce.c Invalid nonce (nonce2 
>> rÍ7B½¿| {Ó9R~R¬FSÖ~Sb~@»êü#025nÿ| 
>> ¹»©Zo#032#YåP~@Ãô#010PA#027#030P@¸ö§£Â\ÆX`#035~MT~\ÞUl~X~Sb~@»êü#025nÿ| 
>> ¹»©Zo#032#YåP~@Ãô#010)
>> Aug 31 13:58:25 openims-kamailio-1 /usr/sbin/kamailio[7824]: DEBUG: 
>> auth [api.c:147]: auth:pre_auth: Invalid nonce value received
>> Aug 31 13:58:25 openims-kamailio-1 /usr/sbin/kamailio[7824]: DEBUG: 
>> auth [api.c:148]: auth:pre_auth: Invalid nonce value received (ret 2)
>>
>> Cordialement/ Best regards
>> Patrice B
>>
>> -----Message d'origine-----
>> De : Daniel-Constantin Mierla [mailto:miconda at gmail.com]
>> Envoyé : vendredi 31 août 2012 10:35
>> À : SIP Router - Kamailio (OpenSER) and SIP Express Router (SER) - 
>> Users Mailing List
>> Cc : BODEVEN Patrice RD-CORE
>> Objet : Re: [SR-Users] Kamailio - Nonce validity between Register and 
>> Invite
>>
>> Hello,
>>
>> On 8/30/12 10:43 AM, patrice.bodeven at orange.com wrote:
>>> Hello,
>>>
>>> Sorry, but i am blocked on the nonce reuse between the Register and 
>>> the INVITE even with the info provided in debug mode.
>>>    My understanding is Nonce should be valid between SIP methods 
>>> until the nonce is valid in the time.
>>> But I don't know how the Nonce is stored and how Kamailio is looking 
>>> for the nonce into the memory !
>>>
>>> I did a test on Register => nonce reuse is ok.
>>> I did a test on INVITE => nonce reuse of Register or on previous 
>>> INVITE doesn't work !
>>>
>>> But based on the site indicated below and the SIPP (generate the 
>>> INVITE with the Nonce of Register), the result is
>>> auth_db [authorize.c:271]: realm value [openims-kamailio-1.mycluster]
>>> auth [api.c:95]: auth: digest-algo: MD5 parsed value: 1
>>> auth [api.c:147]: auth:pre_auth: Invalid nonce value received
>>>
>>> But there is no explanation why the nonce is invalid.
>> can you apply the following patch:
>>
>> http://git.sip-router.org/cgi-bin/gitweb.cgi?p=sip-router;a=blobdiff;f=modules/auth/api.c;h=c79de5c9dd6cd3e176c05d836c7b0a4275f8a3d2;hp=2ee39a7a6cbcce741365dac97ae8a3db8c196b0c;hb=494b383edde7a2d193c220f3117506e4cc95932f;hpb=e4ecf49add0e62330e1db071106806e35e9b078a 
>>
>>
>> It practically adds the return code in the log message, that should give
>> better leads to why the nonce is considered invalid, allowing to track
>> back quicker in the source code.
>>
>> Cheers,
>> Daniel
>>
>


-- 
Daniel-Constantin Mierla - http://www.asipto.com
http://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda
Kamailio Advanced Training, Berlin, Nov 5-8, 2012 - http://asipto.com/u/kat