[pjsip] Panic on symbian

nir elkayam nir.elkayam at gmail.com
Wed Dec 9 17:09:46 EST 2009


hi,

attached call stack that started when I move to 1.5
maybe this is related to the problem,
this happend after I subscribe, I got the first notify and hasn't got yet
the PJSIP_EVSUB_STATE_ACTIVE.

nir

     ntdll.dll!7c90120e()
     ntdll.dll!7c96e139()
     ntdll.dll!7c94b0aa()
     ntdll.dll!7c91a3f5()
     ntdll.dll!7c928854()
     alt.exe!pj_thread_local_set(long index=196608, void *
value=0x003c0178)  Line 813 + 0x10    C
     ntdll.dll!7c96e5df()
     ntdll.dll!7c96eea0()
     ntdll.dll!7c94b394()
     alt.exe!_output(_iobuf * stream=0x00000006, const char *
format=0x010bcf70, char * argptr=0x010bd074)  Line 405 + 0x17    C
     alt.exe!pj_thread_this()  Line 531 + 0xb    C
     ffffff7f()
     alt.exe!_heap_alloc_base(unsigned int size=4036)  Line 212    C
     alt.exe!_heap_alloc_dbg(unsigned int nSize=4000, int nBlockUse=1, const
char * szFileName=0x00000000, int nLine=0)  Line 397 + 0x9    C
     alt.exe!_nh_malloc_dbg(unsigned int nSize=4000, int nhFlag=0, int
nBlockUse=1, const char * szFileName=0x00000000, int nLine=0)  Line 260 +
0x15    C
     alt.exe!malloc(unsigned int nSize=4000)  Line 139 + 0x15    C
     alt.exe!default_block_alloc(pj_pool_factory * factory=0x003c083c,
unsigned int size=4000)  Line 46 + 0x9    C
     alt.exe!pj_pool_create_block(pj_pool_t * pool=0x00b44030, unsigned int
size=4000)  Line 60 + 0x15    C
     alt.exe!pj_pool_allocate_find(pj_pool_t * pool=0x00b44030, unsigned int
size=2000)  Line 139 + 0xd    C
     alt.exe!pj_pool_alloc(pj_pool_t * pool=0x00b44030, unsigned int
size=2000)  Line 62 + 0xd    C
     alt.exe!pjsip_tx_data_encode(pjsip_tx_data * tdata=0x00b44094)  Line
448 + 0x11    C
     alt.exe!mod_on_tx_msg(pjsip_tx_data * tdata=0x00b44094)  Line 609 +
0x9    C
     alt.exe!endpt_on_tx_msg(pjsip_endpoint * endpt=0x003c3e8c,
pjsip_tx_data * tdata=0x00b44094)  Line 976 + 0xc    C
     alt.exe!pjsip_transport_send(pjsip_transport * tr=0x003c0f1c,
pjsip_tx_data * tdata=0x00b44094, const void * addr=0x010bdad8, int
addr_len=16, void * token=0x00b44a74, void (void *, pjsip_tx_data *, long)*
cb=0x004a3b80)  Line 650 + 0x1f    C
     alt.exe!send_response_resolver_cb(int status=0, void *
token=0x00b44a74, const pjsip_server_addresses * addr=0x010bdac8)  Line 1632
+ 0x2a    C
     alt.exe!pjsip_resolve(pjsip_resolver_t * resolver=0x003ce848, pj_pool_t
* pool=0x00b44030, const pjsip_host_info * target=0x00b4293c, void *
token=0x00b44a74, void (int, void *, const pjsip_server_addresses *)*
cb=0x004a3c60)  Line 292 + 0x17    C
     alt.exe!pjsip_endpt_resolve(pjsip_endpoint * endpt=0x003c3e8c,
pj_pool_t * pool=0x00b44030, pjsip_host_info * target=0x00b4293c, void *
token=0x00b44a74, void (int, void *, const pjsip_server_addresses *)*
cb=0x004a3c60)  Line 1043 + 0x1c    C
     alt.exe!pjsip_endpt_send_response(pjsip_endpoint * endpt=0x003c3e8c,
pjsip_response_addr * res_addr=0x00b42918, pjsip_tx_data * tdata=0x00b44094,
void * token=0x00b42874, void (pjsip_send_state *, long, int *)*
cb=0x004ad6e0)  Line 1689 + 0x20    C
     alt.exe!tsx_send_msg(pjsip_transaction * tsx=0x00b42874, pjsip_tx_data
* tdata=0x00b44094)  Line 1929 + 0x23    C
     alt.exe!tsx_on_state_proceeding_uas(pjsip_transaction * tsx=0x00b42874,
pjsip_event * event=0x010be324)  Line 2357 + 0xd    C
     alt.exe!tsx_on_state_trying(pjsip_transaction * tsx=0x00b42874,
pjsip_event * event=0x010be324)  Line 2287 + 0xd    C
     alt.exe!pjsip_tsx_send_msg(pjsip_transaction * tsx=0x00b42874,
pjsip_tx_data * tdata=0x00b44094)  Line 1590 + 0x10    C
     alt.exe!pjsip_dlg_send_response(pjsip_dialog * dlg=0x00b3d474,
pjsip_transaction * tsx=0x00b42874, pjsip_tx_data * tdata=0x00b44094)  Line
1415 + 0xd    C
     alt.exe!on_tsx_state_uac(pjsip_evsub * sub=0x00b3dc64,
pjsip_transaction * tsx=0x00b42874, pjsip_event * event=0x010beaac)  Line
1848 + 0x17    C
     alt.exe!mod_evsub_on_tsx_state(pjsip_transaction * tsx=0x00b42874,
pjsip_event * event=0x010beaac)  Line 2077 + 0x11    C
     alt.exe!pjsip_dlg_on_tsx_state(pjsip_dialog * dlg=0x00b3d474,
pjsip_transaction * tsx=0x00b42874, pjsip_event * e=0x010beaac)  Line 1918 +
0x1a    C
     alt.exe!mod_ua_on_tsx_state(pjsip_transaction * tsx=0x00b42874,
pjsip_event * e=0x010beaac)  Line 178 + 0x11    C
     alt.exe!tsx_set_state(pjsip_transaction * tsx=0x00b42874,
pjsip_tsx_state_e state=PJSIP_TSX_STATE_TRYING, pjsip_event_id_e
event_src_type=PJSIP_EVENT_RX_MSG, void * event_src=0x003cec0c)  Line 1108 +
0x13    C
     alt.exe!tsx_on_state_null(pjsip_transaction * tsx=0x00b42874,
pjsip_event * event=0x010bece0)  Line 2071 + 0x14    C
     alt.exe!pjsip_tsx_recv_msg(pjsip_transaction * tsx=0x00b42874,
pjsip_rx_data * rdata=0x003cec0c)  Line 1627 + 0x10    C
     alt.exe!pjsip_dlg_on_rx_request(pjsip_dialog * dlg=0x00b3d474,
pjsip_rx_data * rdata=0x003cec0c)  Line 1593 + 0xd    C
     alt.exe!mod_ua_on_rx_request(pjsip_rx_data * rdata=0x003cec0c)  Line
683 + 0xd    C
     alt.exe!endpt_on_rx_msg(pjsip_endpoint * endpt=0x003c3e8c, int
status=0, pjsip_rx_data * rdata=0x003cec0c)  Line 904 + 0xf    C
     alt.exe!pjsip_tpmgr_receive_packet(pjsip_tpmgr * mgr=0x003ce6d8,
pjsip_rx_data * rdata=0x003cec0c)  Line 1473 + 0x15    C
     alt.exe!udp_on_read_complete(pj_ioqueue_key_t * key=0x003cc33c,
pj_ioqueue_op_key_t * op_key=0x003cec18, long bytes_read=1103)  Line 166 +
0x16    C
     alt.exe!ioqueue_dispatch_read_event(pj_ioqueue_t * ioqueue=0x003cbd88,
pj_ioqueue_key_t * h=0x003cc33c)  Line 552 + 0x14    C
     alt.exe!pj_ioqueue_poll(pj_ioqueue_t * ioqueue=0x003cbd88, const
pj_time_val * timeout=0x010bfc0c)  Line 765 + 0x17    C
     alt.exe!pjsip_endpt_handle_events2(pjsip_endpoint * endpt=0x003c3e8c,
const pj_time_val * max_timeout=0x010bfdd8, unsigned int *
p_count=0x00000000)  Line 719 + 0x10    C
>    alt.exe!pjsip_endpt_handle_events(pjsip_endpoint * endpt=0x003c3e8c,
const pj_time_val * max_timeout=0x010bfdd8)  Line 747 + 0xf    C
     alt.exe!SipEngine::process()  Line 395 + 0x13    C++
     alt.exe!SipEngine::worker_thread(void * arg=0x00000000)  Line 116
C++
     alt.exe!thread_main(void * param=0x003c0b28)  Line 428 + 0xf    C
     kernel32.dll!7c80b729()


On Wed, Dec 9, 2009 at 4:37 PM, nir elkayam <nir.elkayam at gmail.com> wrote:

> hi,
>
> I don't think its network related as it usually happen when i subscribe
> just after registration so i don't think its network related.
>
> at first i thoght that its relate to having several subscribe at once so I
> do them one by one and use the notify to drive this process.
>
> I'll recheck again without this "feature" to see where it crushes more,
> maybe this will give some hist into the problem.
>
> nir
>
>
> On Wed, Dec 9, 2009 at 4:17 PM, Benny Prijono <bennylp at teluu.com> wrote:
>
>> The port difference was another problem, as we discussed in the other
>> thread.
>>
>> Re: the panic, this probably is the same problem as
>> http://trac.pjsip.org/repos/ticket/1002 which I just added, and was
>> discussed on another thread. I'm now working on this, and I can see
>> few problems (read: crashes :) already, but these problems only
>> manifest themselves if I disrupt my network connectivity (i.e.
>> enabling/disabling network adapter) while the test program is running
>> continuously in the background.
>>
>> So in your cases, do you think you have a problem with network
>> connectivity during your tests? Just trying to narrow down the issue.
>>
>> Cheers
>>  Benny
>>
>> On Mon, Dec 7, 2009 at 1:08 PM, nir elkayam <nir.elkayam at gmail.com>
>> wrote:
>> > I sent a mail the other day about a problem I have in unsubscribe. when
>> I
>> > send the unsubscribe, it send it to port other then the one needed
>> (30100
>> > instade of 30101).
>> >
>> > maybe there is some memory override in subscribe?
>> >
>> > nir
>> >
>> > On Mon, Dec 7, 2009 at 3:03 PM, Vasyl Strutynskyy
>> > <vasyl.strutynskyy at gmail.com> wrote:
>> >>
>> >> Nir,
>> >>
>> >> We faced with exactly the same problem. And I'm sure this bug is inside
>> >> PJSIP, where exactly I don't know. But it is definitely related with
>> >> SUBSCRIBE.
>> >>
>> >> Unfortunately we didn't receive fix for SUBSCRIBE, but one workaround
>> was
>> >> applicable for us, where we changed subscribing to unsolicited MWI
>> events.
>> >>
>> >> You can see changed provided for unsolicited MWI here:
>> >> http://trac.pjsip.org/repos/changeset/3021
>> >>
>> >> Hence I think you should try to ask PJSIP to fix this bug with
>> SUBSCRIBE
>> >> or try to fix yourself. Just analyze code related with SUBSCRIBE.
>> >>
>> >> Best Regards
>> >> *Vasyl
>> >>
>> >> 2009/12/7 nir elkayam <nir.elkayam at gmail.com>
>> >>>
>> >>> Hi Vasyl (and all list too..)
>> >>>
>> >>> I have just few hints about the panic, but not getting it always. It
>> >>> happend more on my N95, It almost never happend on N81 and 5800 (which
>> I use
>> >>> to test).
>> >>>
>> >>> It happen most when I start pjsua, immediatly register, and on the
>> >>> callback on secsessful registration i start the subscribe. If i
>> subscribre
>> >>> only later, it goes on without panic. also the panic doesn't occur all
>> the
>> >>> times.
>> >>>
>> >>> I added flush to the logging to file on both and on the panic it
>> always
>> >>> in the middle of writing some sip msg to the log file.
>> >>>
>> >>> I suspect of long msg/long log msg, but when disabled the log to file
>> it
>> >>> also happend, so maybe some place else,  ?!
>> >>>
>> >>> nir
>> >>>
>> >>> On Mon, Dec 7, 2009 at 2:32 PM, Vasyl Strutynskyy
>> >>> <vasyl.strutynskyy at gmail.com> wrote:
>> >>>>
>> >>>> Hi, nir
>> >>>>
>> >>>> Can you specify use case to reproduce it in details.
>> >>>>
>> >>>> Actually we faced with one KERN EXEC 3 using SUBSCRIBE command to
>> >>>> receive MWI notifications. And now it is known fact for us that
>> SUBSCRIBE
>> >>>> provoke crash.
>> >>>>
>> >>>> Best Regards
>> >>>> *Vasyl
>> >>>>
>> >>>> 2009/12/7 nir elkayam <nir.elkayam at gmail.com>
>> >>>>>
>> >>>>> hi,
>> >>>>>
>> >>>>> I am facing Panic (KERN - EXEC 3) using pjsip on symbian.
>> >>>>> It started when I began using the buddies and when no buddies, no
>> >>>>> panic. I am checking my sw but, has anyone had problems and panic
>> like that
>> >>>>> relate to pjsip errors/mishandeling?
>> >>>>>
>> >>>>> thanks,
>> >>>>> nir
>> >>>>>
>> >>>>>
>> >>>>> _______________________________________________
>> >>>>> Visit our blog: http://blog.pjsip.org
>> >>>>>
>> >>>>> pjsip mailing list
>> >>>>> pjsip at lists.pjsip.org
>> >>>>> http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org
>> >>>>>
>> >>>>
>> >>>>
>> >>>> _______________________________________________
>> >>>> Visit our blog: http://blog.pjsip.org
>> >>>>
>> >>>> pjsip mailing list
>> >>>> pjsip at lists.pjsip.org
>> >>>> http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org
>> >>>>
>> >>>
>> >>>
>> >>>
>> >>> --
>> >>> ניר אלקיים
>> >>> טל: 050-3930056
>> >>> nir.elkayam at gmail.com
>> >>>
>> >>>
>> >>> _______________________________________________
>> >>> Visit our blog: http://blog.pjsip.org
>> >>>
>> >>> pjsip mailing list
>> >>> pjsip at lists.pjsip.org
>> >>> http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org
>> >>>
>> >>
>> >>
>> >> _______________________________________________
>> >> Visit our blog: http://blog.pjsip.org
>> >>
>> >> pjsip mailing list
>> >> pjsip at lists.pjsip.org
>> >> http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org
>> >>
>> >
>> >
>> >
>> > --
>> > ניר אלקיים
>> > טל: 050-3930056
>> > nir.elkayam at gmail.com
>> >
>> >
>> > _______________________________________________
>> > Visit our blog: http://blog.pjsip.org
>> >
>> > pjsip mailing list
>> > pjsip at lists.pjsip.org
>> > http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org
>> >
>> >
>>
>>
>>
>> --
>> Best regards,
>>
>>  Benny
>>
>> _______________________________________________
>> Visit our blog: http://blog.pjsip.org
>>
>> pjsip mailing list
>> pjsip at lists.pjsip.org
>> http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org
>>
>
>
>
> --
> ניר אלקיים
> טל: 050-3930056
> nir.elkayam at gmail.com
>
>


-- 
ניר אלקיים
טל: 050-3930056
nir.elkayam at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.pjsip.org/pipermail/pjsip_lists.pjsip.org/attachments/20091210/11032e75/attachment.html>


More information about the pjsip mailing list