[SR-Users] CRITICAL ERROR in data_lamp.c!

[Daniel-Constantin Mierla]

What crash with htable api? I can't see anything related in this discussion.

On 11.06.19 17:15, Mojtaba wrote:
> Thank you.
> And what about the crash when using it with htable api module?
>
> On Tue, 11 Jun 2019, 18:18 Daniel-Constantin Mierla <miconda at gmail.com
> <mailto:miconda at gmail.com> wrote:
>
>     Yes, you must allocate with pkg, because at the end those chunks
>     are freed, once the new message is built. You can also look at how
>     other operations with data lumps are done inside the textops module.
>
>     On 11.06.19 14:44, Mojtaba wrote:
>>     I assigned a const sdp string to it just for checking. Does it
>>     matter to allocate new memory for newbody?
>>
>>     On Tue, 11 Jun 2019, 15:22 Daniel-Constantin Mierla
>>     <miconda at gmail.com <mailto:miconda at gmail.com> wrote:
>>
>>         Hello,
>>
>>         how do you set newbody.s and newbody.len?
>>
>>         Daniel
>>
>>         On 11.06.19 12:29, Mojtaba wrote:
>>         > Not yet, because it is under developing, i could paste it
>>         here right away.
>>         >
>>         > {
>>         >  .
>>         >  .
>>         >  .
>>         >     body.s = get_body(msg);
>>         >     if (body.s == 0) {
>>         >                 LM_ERR("failed to get the message body\n");
>>         >         return -1;
>>         >     }
>>         >     body.len = msg->len - (int) (body.s - msg->buf);
>>         > //    body.len = strlen(body.s);
>>         >     if (body.len == 0) {
>>         >                 LM_DBG("message body has zero length\n");
>>         >         return -1;
>>         >     }
>>         >     l = del_lump(msg, body.s - msg->buf, body.len, 0);
>>         >     if (!l) {
>>         >                 LM_ERR("del_lump failed\n");
>>         >         return -1;
>>         >     }
>>         >
>>         >     if (!insert_new_lump_after(l, newbody.s, newbody.len, 0)) {
>>         >                 LM_ERR("insert_new_lump_after failed\n");
>>         >         return -1;
>>         >     }
>>         > }
>>         > I asked related issue in this regards in Kamailio developer
>>         forum,
>>         > before. let me paste it at below. The means of
>>         other_process function
>>         > is above code.
>>         > In during developing new module in Kamailio, I encountered with
>>         > strange behaviour. I used htable module instead of linked
>>         list for
>>         > save some temporary data during call process.dependencies
>>         module to
>>         > save some information. the skeleton of my code is like thus:
>>         >
>>         > if (SIP_REQUEST &METHOD_INVITE) {
>>         >           insert_data_htable("some data");
>>         >           other_process(msg);
>>         > }else if(SIP_REPLY){
>>         >          get_data_htable("some data");
>>         >          other_process(msg);
>>         > }
>>         > In other_process function i have some changes in msg's body
>>         with message lump.
>>         > But after i make first call in kamailio. The kamailio was
>>         crashed.What
>>         > is it's problem?
>>         > In additional if i comment either insert/get_data_htable
>>         function or
>>         > othet_process function, i dont have this issue.
>>         > With best regards.
>>         >
>>         >
>>         > On Tue, Jun 11, 2019 at 2:46 PM Daniel-Constantin Mierla
>>         > <miconda at gmail.com <mailto:miconda at gmail.com>> wrote:
>>         >> Hello,
>>         >>
>>         >> likely you didn't allocate the pointer inserted in data
>>         lump list. Is
>>         >> the code using data lumps available to look at on git?
>>         >>
>>         >> Cheers,
>>         >> Daniel
>>         >>
>>         >> On 11.06.19 12:01, Mojtaba wrote:
>>         >>> Hello,
>>         >>> What does this CRITICAL error mean?
>>         >>> Jun 11 05:48:13 debian8 /usr/local/sbin/kamailio[8468]:
>>         {1 1 INVITE
>>         >>> Njg0MzNhMzJhOTY3MzQ3NmNlNWY0ODkzMTJmNTM1ZTk.} CRITICAL:
>>         <core>
>>         >>> [core/mem/q_malloc.c:486]: qm_free(): BUG: bad pointer
>>         0xa3c188 (out
>>         >>> of memory block!) called from core: core/data_lump.c:
>>         free_lump(466) -
>>         >>> aborting
>>         >>>
>>         >>> I got it when i make new call during debugging new module
>>         in Kamailio?
>>         >>> In module we have to change some message's body with
>>         data_lump.
>>         >>> I asked relaed issue (strange behaviour while developing
>>         new module)
>>         >>> in Kamailio developer forum <sr-dev at lists.kamailio.org
>>         <mailto:sr-dev at lists.kamailio.org>>
>>         >>> With Best Regards.
>>         >>>
>>         >> --
>>         >> Daniel-Constantin Mierla -- www.asipto.com
>>         <http://www.asipto.com>
>>         >> www.twitter.com/miconda <http://www.twitter.com/miconda>
>>         -- www.linkedin.com/in/miconda
>>         <http://www.linkedin.com/in/miconda>
>>         >>
>>         >
>>         > --
>>         > --Mojtaba Esfandiari.S
>>
>>         -- 
>>         Daniel-Constantin Mierla -- www.asipto.com
>>         <http://www.asipto.com>
>>         www.twitter.com/miconda <http://www.twitter.com/miconda> --
>>         www.linkedin.com/in/miconda <http://www.linkedin.com/in/miconda>
>>
>     -- 
>     Daniel-Constantin Mierla -- www.asipto.com <http://www.asipto.com>
>     www.twitter.com/miconda <http://www.twitter.com/miconda> -- www.linkedin.com/in/miconda <http://www.linkedin.com/in/miconda>
>
-- 
Daniel-Constantin Mierla -- www.asipto.com
www.twitter.com/miconda -- www.linkedin.com/in/miconda

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.kamailio.org/pipermail/sr-users/attachments/20190612/49ef44f2/attachment.html>