[sr-dev] crash at f_malloc.c

[Daniel-Constantin Mierla]

I just pushed a patch, no need to send the backtraces anymore, try with
latest master to see if there is any issue.

Daniel

On 20/09/15 22:20, Daniel-Constantin Mierla wrote:
> These two have different backtraces, so the previous commands won't work
> because of variables not being present in the frame. Send first 'bt' for
> each of the cores.
>
> Daniel
>
> On 20/09/15 20:31, Juha Heinanen wrote:
>> i built new kamailio from latest master and got again the crash.  this time
>> it produced two core files:
>>
>> 45852 -rw------- 1 root root 59764736 Sep 20 21:21 /var/cores/core.sip-proxy.sig11.32373
>> 45856 -rw------- 1 root root 59805696 Sep 20 21:21 /var/cores/core.sip-proxy.sig11.32406
>>
>> /var/cores/core.sip-proxy.sig11.32373 gives:
>>
>> (gdb) frame 0
>> #0  fm_free (qmp=0xa42d00 <_shm_root>, p=0x2d64726f6365520a) at mem/f_malloc.c:589
>> 589	in mem/f_malloc.c
>> (gdb) list
>> 584	in mem/f_malloc.c
>> (gdb) info locals
>> f = 0x7f78ef610780
>> __FUNCTION__ = "fm_free"
>> (gdb) p *qm
>> value has been optimized out
>> (gdb) p *frag
>> No symbol "frag" in current context.
>> (gdb) p *n
>> No symbol "n" in current context.
>>
>> and /var/cores/core.sip-proxy.sig11.32406 gives:
>>
>> (gdb) frame 0
>> #0  fm_extract_free (frag=<optimized out>, qm=<optimized out>) at mem/f_malloc.c:181
>> 181	in mem/f_malloc.c
>> (gdb) list
>> 176	in mem/f_malloc.c
>> (gdb) info locals
>> hash = 2
>> (gdb) p *qm
>> value has been optimized out
>> (gdb) p *frag
>> value has been optimized out
>> (gdb) p *n
>> No symbol "n" in current context.
>>
>> -- juha

-- 
Daniel-Constantin Mierla
http://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda
Book: SIP Routing With Kamailio - http://www.asipto.com
Kamailio Advanced Training, Sep 28-30, 2015, in Berlin - http://asipto.com/u/kat