Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Daniel-Constantin Mierla
On 03/05/16 17:31, Juha Heinanen wrote: > Daniel-Constantin Mierla writes: > >> And I think I found the issue -- when the type was not provided, the >> pointer inside the xmlNode structure was used instead of using the >> function to return a clone of the content, in order to free it later via >>

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Juha Heinanen
Daniel-Constantin Mierla writes: > And I think I found the issue -- when the type was not provided, the > pointer inside the xmlNode structure was used instead of using the > function to return a clone of the content, in order to free it later via > xmlrpc module garbage collector. > > I couldn't

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Daniel-Constantin Mierla
Thanks for digging in further, the details proved to be important. And I think I found the issue -- when the type was not provided, the pointer inside the xmlNode structure was used instead of using the function to return a clone of the content, in order to free it later via xmlrpc module garbage

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Juha Heinanen
in my sems xmlrpc test, i changed htable.reload command to permissions.trustedReload (that does not have parameters) and have not been able to produce a crash. my conclusion is that xmlrpc string parameter value without is causing the crash (perhaps due to memory allocation bug either in kamailio

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Juha Heinanen
string parameter value without tags is legal in xmlrpc: Scalar s s can be scalars, type is indicated by nesting the value inside one of the tags listed in this table: Tag TypeExample orfour-byte signed integer-12 0 (false) or 1 (true) 1 string hello world

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Juha Heinanen
Daniel-Constantin Mierla writes: > I checked the xmlrpc module commit log and last code changes were in > September 2014 - even those are not related to any xml ops. In the past, > have you used the xmlrpc commands for other similar deployments with an > older version than 4.4 of kamailio? I have

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Daniel-Constantin Mierla
Still the issue is in the system memory, even the xmlrpc module was compiled with pkg malloc manager. The crash this time is in regexp functions from libc. It could be the operations done with libxml2 inside the xmlrpc, because they use the libxml2 native functions that rely on system malloc. I

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Juha Heinanen
Daniel-Constantin Mierla writes: > Thinking more of it -- is the route[cleaner] executed with rtimer on a > dedicated timer process? I have two timer routes: modparam("rtimer", "timer", "name=stat_timer;interval=60;mode=1") modparam("rtimer", "exec", "timer=stat_timer;route=stats") modparam("rti

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Juha Heinanen
Daniel-Constantin Mierla writes: > What is the version of kamailio running there? 4.4 build from git on April 7. > xmlrpc is compiled by default to use system memory manager (expecting to > need big chunks for some responses), can you recompile it to use the > pkg? You have to edit modules/xmlrp

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Daniel-Constantin Mierla
On 03/05/16 09:01, Juha Heinanen wrote: > Daniel-Constantin Mierla writes: > >> These ones are no longer related to xmlrpc, but triggered by an sqlops >> query - first bracktrace, the other two are during the signal >> handling. > OK, but there must be some kind of relation to xmlrpc, because I h

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Daniel-Constantin Mierla
On 03/05/16 09:01, Juha Heinanen wrote: > Daniel-Constantin Mierla writes: > >> These ones are no longer related to xmlrpc, but triggered by an sqlops >> query - first bracktrace, the other two are during the signal >> handling. > OK, but there must be some kind of relation to xmlrpc, because I h

Re: [SR-Users] xmlrpc crash

2016-05-03 Thread Juha Heinanen
Daniel-Constantin Mierla writes: > These ones are no longer related to xmlrpc, but triggered by an sqlops > query - first bracktrace, the other two are during the signal > handling. OK, but there must be some kind of relation to xmlrpc, because I have had this same test system running without any

Re: [SR-Users] xmlrpc crash

2016-05-02 Thread Daniel-Constantin Mierla
These ones are no longer related to xmlrpc, but triggered by an sqlops query - first bracktrace, the other two are during the signal handling. The one for sqlops leads quite a bit inside the libmysqlclient, which can happen because there is a buffer overflow in kamailio for the operations in the s

Re: [SR-Users] xmlrpc crash

2016-05-02 Thread Juha Heinanen
Daniel-Constantin Mierla writes: > do you get any error messages in the logs or error code in the xmlrpc > reply? no error messages. the crash does not happen immediate when xmlrpc request is served. in the example below, the crash came about 20 sec later perhaps as result of some garbage colle

Re: [SR-Users] xmlrpc crash

2016-05-02 Thread Daniel-Constantin Mierla
Hello, do you get any error messages in the logs or error code in the xmlrpc reply? In frame 2 the msg structure is NULL, should be the associated sip-http request. Cheers, Daniel On 01/05/16 15:48, Juha Heinanen wrote: > i was sending some xmlrpc requests to kamailio and got crash. > > -- juha

[SR-Users] xmlrpc crash

2016-05-01 Thread Juha Heinanen
i was sending some xmlrpc requests to kamailio and got crash. -- juha (gdb) where #0 0x7f9c465da634 in __GI___libc_free (mem=0x17828c0) at malloc.c:2945 #1 0x7f9c39e0b797 in collect_garbage () at xmlrpc.c:741 #2 0x7f9c39e12669 in dispatch_rpc (msg=0x0, s1=0x17828b0 "ams", s2=0x7f9