Re: PJSUA crash when trying to cancel outgoing call with pjsua_call_answer();

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi,

Yes, so things like --disable-ssl and --enable-ssl only works in the --disable-ssl direction, setting --enable-ssl actually has the same effect as --disable-ssl in that it disables ssl.

Same with --{enable,disable}-webrtc.

As a packager myself I want to explicitly specify each and every enable/disable.  So if I give --enable-X I want ./configure to FAIL if it can't enable X.  Current status of this is just a nightmare.  Current status is that I need to not specify --enable-X and then HOPE that it does the right thing.


Kind Regards,
Jaco Kroon

On 2019/12/04 11:33, Skale Franz wrote:

Hi Jaco,
it's not good to be ignored.
I found a problem when it comes to packaging eg. on DEBIAN.
When using the external-lib configure flags, and even installing the dependecies, PJPROJECT won't work when not compiled with third party libs.
This is for the version 2.7.2 as well as 2.9.
They should give us the option to raise issues on github IMHO.

Rgds.
Franz

________________________________________
Von: pjsip <pjsip-bounces@xxxxxxxxxxxxxxx> im Auftrag von Jaco Kroon <jaco@xxxxxxxxx>
Gesendet: Mittwoch, 04. Dezember 2019 10:28
An: pjsip@xxxxxxxxxxxxxxx
Betreff: Re:  PJSUA crash when trying to cancel outgoing call with pjsua_call_answer();

Hi Franz,

You're not the only person with the issue of being unable to log issues.

I've reported ./configure issues via github, was told to send it to the
ML ... sent it to the ML, got ignored.

So I don't know ... just carrying patches locally now.  Which is a pain.

Kind Regards,
Jaco

On 2019/12/04 11:21, Skale Franz wrote:
Hi Andreas,
i'm asking myself where to report bugs anyway.
You cannot push trac issues, neither the github repo allows you to create incidents !
This is a problem, especially when it comes to bugs and you hit one !

Line starting with 1763 is of special interest:
You transaction is in progress so it leaved the function
/* Can't terminate dialog because transaction is in progress.
    pjsip_dlg_terminate(dlg);
     */
    goto on_return;

But your program tries to pushback an uninitilaized list which then aborts on line 2638:
pj_list_push_back(&call->async_call.call_var.inc_call.answers, answer);

It's not clear to me why the code doesn't catch the goto line in the first place.
It would be great, if a core developer would respond anyway.

Rgds.
Franz

Von: pjsip <pjsip-bounces@xxxxxxxxxxxxxxx> im Auftrag von Andreas Wehrmann <a.wehrmann@xxxxxxxxxx>
Gesendet: Dienstag, 03. Dezember 2019 16:42
An: pjsip@xxxxxxxxxxxxxxx
Betreff:  PJSUA crash when trying to cancel outgoing call with pjsua_call_answer();

Hey folks,

By accident I found a (reproducible) crash in PJSUA.
So I found this because I was using the wrong PJSUA call for cancelling
an outgoing call.
The scenario is this:

- Initiate call with pjsua_call_make_call()
- Before the call is picked up; try to cancel the call with
pjsua_call_answer2() (and providing msg_data).

In this case, PJSUA always crashes with the following backtrace:

#0  0x0000000000571de0 in pj_list_insert_after (pos=0x7fffe0005608,
node=0x7fffe00088a8) at ../include/pj/list_i.h:33
#1  0x0000000000571e18 in pj_list_insert_before (pos=0x83fc28
<pjsua_var+68200>, node=0x7fffe00088a8) at ../include/pj/list_i.h:40
#2  0x00000000004581f7 in pj_list_push_back (list=0x83fc28
<pjsua_var+68200>, node=0x7fffe00088a8) at ../../pjlib/include/pj/list.h:128
#3  0x000000000045dad8 in pjsua_call_answer2 (call_id=1, opt=0x0,
code=603, reason=0x0, msg_data=0x7ffff0854870) at
../src/pjsua-lib/pjsua_call.c:2638
#4  0x000000000042abaf in siss::UserAgent::answer (this=0x8c61c0,
cid=..., status_code=status_code@entry=603, reason=...,
additional_header=std::vector of length 0, capacity 0,
additional_payload=std::vector of length 0, capacity 0)
     at useragent.cpp:1071
#5  0x000000000042ae18 in siss::UserAgent::hangup (this=<optimized out>,
cid=..., status_code=status_code@entry=603, reason=...) at
useragent.cpp:1093


Now before you get angry: Yes, I realize I need to use
pjsua_call_hangup() instead (and already confirmed it works that way);
But I would have expected the function to return with the appropiate
error code and not cause a hard crash.

Best Regards,
Andreas Wehrmann


_______________________________________________
Visit our blog: http://blog.pjsip.org

pjsip mailing list
pjsip@xxxxxxxxxxxxxxx
http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org

_______________________________________________
Visit our blog: http://blog.pjsip.org

pjsip mailing list
pjsip@xxxxxxxxxxxxxxx
http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org
_______________________________________________
Visit our blog: http://blog.pjsip.org

pjsip mailing list
pjsip@xxxxxxxxxxxxxxx
http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org

_______________________________________________
Visit our blog: http://blog.pjsip.org

pjsip mailing list
pjsip@xxxxxxxxxxxxxxx
http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org
_______________________________________________
Visit our blog: http://blog.pjsip.org

pjsip mailing list
pjsip@xxxxxxxxxxxxxxx
http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org

[Index of Archives]     [Asterisk Users]     [Asterisk App Development]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Security]     [Bugtraq]     [Linux]     [Linux OMAP]     [Linux MIPS]     [Linux API]
  Powered by Linux