[asterisk-bugs] [JIRA] (ASTERISK-26946) Crash at the user leaves the conference when the video frame is distributed

[Richard Mudgett (JIRA)]

     [ https://issues.asterisk.org/jira/browse/ASTERISK-26946?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Richard Mudgett updated ASTERISK-26946:
---------------------------------------

    Description: 
when the src distribution the video frame to others in the confbridge，another user leave，it will crash。（join and leave confbridge was
 very frequently use sipp）
the crash back trace was like that：
{noformat}
(gdb) bt
#0  ast_bridge_queue_everyone_else (bridge=<value optimized out>, bridge_channel=0x1dec7a8, frame=0x7f8f58c619f0) at bridge_channel.c:1108
#1  0x00007f8f6e05a2cc in softmix_bridge_write_video (bridge=0x1e3fb28, bridge_channel=<value optimized out>, frame=0x7f8f58c619f0)
    at bridge_softmix.c:561
#2  softmix_bridge_write (bridge=0x1e3fb28, bridge_channel=<value optimized out>, frame=0x7f8f58c619f0) at bridge_softmix.c:705
#3  0x00007f8f6d42b127 in send_pic (data=0x675efc8) at app_confbridge.c:9084
#4  0x0000000000591bad in ast_sched_runq (con=0x7f8f4748e230) at sched.c:625
#5  0x00000000005920c1 in sched_run (data=0x7f8f4748e230) at sched.c:140
#6  0x00000000005c46ab in dummy_start (data=<value optimized out>) at utils.c:1232
#7  0x0000003b210079d1 in start_thread () from /lib64/libpthread.so.0
#8  0x0000003b20ce8b6d in clone () from /lib64/libc.so.6
{noformat}

BTW，i want know why the bridge->channels didn't have the lock,it will operate by muti threads,that's Make me wonder.

  was:
when the src distribution the video frame to others in the confbridge，another user leave，it will crash。（join and leave confbridge was
 very frequently use sipp）
the crash back trace was like that：
(gdb) bt
#0  ast_bridge_queue_everyone_else (bridge=<value optimized out>, bridge_channel=0x1dec7a8, frame=0x7f8f58c619f0) at bridge_channel.c:1108
#1  0x00007f8f6e05a2cc in softmix_bridge_write_video (bridge=0x1e3fb28, bridge_channel=<value optimized out>, frame=0x7f8f58c619f0)
    at bridge_softmix.c:561
#2  softmix_bridge_write (bridge=0x1e3fb28, bridge_channel=<value optimized out>, frame=0x7f8f58c619f0) at bridge_softmix.c:705
#3  0x00007f8f6d42b127 in send_pic (data=0x675efc8) at app_confbridge.c:9084
#4  0x0000000000591bad in ast_sched_runq (con=0x7f8f4748e230) at sched.c:625
#5  0x00000000005920c1 in sched_run (data=0x7f8f4748e230) at sched.c:140
#6  0x00000000005c46ab in dummy_start (data=<value optimized out>) at utils.c:1232
#7  0x0000003b210079d1 in start_thread () from /lib64/libpthread.so.0
#8  0x0000003b20ce8b6d in clone () from /lib64/libc.so.6


BTW，i want know why the bridge->channels didn't have the lock,it will operate by muti threads,that's Make me wonder.


> Crash at the user leaves the conference when the video frame is distributed
> ---------------------------------------------------------------------------
>
>                 Key: ASTERISK-26946
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-26946
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: . I did not set the category correctly.
>    Affects Versions: 13.1.0
>         Environment: centOS 6.5
>            Reporter: junxu
>
> when the src distribution the video frame to others in the confbridge，another user leave，it will crash。（join and leave confbridge was
>  very frequently use sipp）
> the crash back trace was like that：
> {noformat}
> (gdb) bt
> #0  ast_bridge_queue_everyone_else (bridge=<value optimized out>, bridge_channel=0x1dec7a8, frame=0x7f8f58c619f0) at bridge_channel.c:1108
> #1  0x00007f8f6e05a2cc in softmix_bridge_write_video (bridge=0x1e3fb28, bridge_channel=<value optimized out>, frame=0x7f8f58c619f0)
>     at bridge_softmix.c:561
> #2  softmix_bridge_write (bridge=0x1e3fb28, bridge_channel=<value optimized out>, frame=0x7f8f58c619f0) at bridge_softmix.c:705
> #3  0x00007f8f6d42b127 in send_pic (data=0x675efc8) at app_confbridge.c:9084
> #4  0x0000000000591bad in ast_sched_runq (con=0x7f8f4748e230) at sched.c:625
> #5  0x00000000005920c1 in sched_run (data=0x7f8f4748e230) at sched.c:140
> #6  0x00000000005c46ab in dummy_start (data=<value optimized out>) at utils.c:1232
> #7  0x0000003b210079d1 in start_thread () from /lib64/libpthread.so.0
> #8  0x0000003b20ce8b6d in clone () from /lib64/libc.so.6
> {noformat}
> BTW，i want know why the bridge->channels didn't have the lock,it will operate by muti threads,that's Make me wonder.



--
This message was sent by Atlassian JIRA
(v6.2#6252)