asterisk users - Mar 2005 - Realtime Problem = Segmentation faults

Hi:
    I had asterisk with RealTime database working perfectly in a RH 9.0 
machine.  I used the sip cache so I even had MWI working.  The problem 
is that I decided to move to Fedora Core 3.  I installed the lastets cvs 
version of asterisk and the RealTime addon from asterisk-addons.  I at 
first had the problems with the kernel and the zaptel driver but all 
that was solved with the configuration from the Asterisk Wiki.  Then 
when I moved my configuration to the new asterisk server and configured 
the RealTime addon it falls in a Segmentation fault.  If I do not load 
the res_config_mysql.so (edited at modules.conf) then asterisks runs 
without any problem.  But if I load the module from boot or from the 
asterisk command load res_config_mysql.so then I get the Segmentation 
fault again.

I'm not sure what the problem is.  Is it a Fedora Core 3 problem, or an 
Asterisk latest version problem? 

I don't think it is a configuration problem because I just used the same 
configuration I had before.  The only diferences may be the OS and 
probably the asterisk version that is only one week newer than the one I 
was running in the old asterisk server, so I'm probably even running the 
same version of asterisk in both machines.

Any advise?  Someone else have a similar configuration working with 
Fedora Core 3?

Thanks in advance,

-- 
Jose R. Ortiz Ubarri (CHEO), CS
System Administrator / Programmer
High Performance Computing facility - UPR
Email: Jose.Ortiz@hpcf.upr.edu|cheo@hpcf.upr.edu
Phone: 787-758-3054
Fax: 787-758-3058

Jose R. Ortiz Ubarri wrote:
>  But if I
> load the module from boot or from the asterisk command load
> res_config_mysql.so then I get the Segmentation fault again.
Where is your backtrace? I don't see a backtrace anywhere.

"Hi. My phone isn't working but I'm not going to let you see what I
did to
cause it to stop working."

Send backtrace from the core file when it crashes.

-Matthew

Matthew:
    I did the cvs checkout asterisk today.  I think I have the latest 
version:
The trace is:
(gdb) backtrace
#0  0x007642b8 in strcasecmp () from /lib/tls/libc.so.6
#1  0xf6eb58c0 in build_peer (name=0x0, v=0x92a42d0, realtime=0) at 
chan_sip.c:9255
#2  0xf6eb67b0 in find_peer (peer=0x0, sin=0x929bf4c, realtime=1) at 
chan_sip.c:1222
#3  0xf6ebea77 in check_user_full (p=0x929bdf0, req=0xf6e9bb50, 
cmd=0xf6e9bd64 "SUBSCRIBE", uri=0xf6e9bd6e
"sip:1002@136.145.54.70:5060",
    reliable=0, sin=0xf6e9bb40, ignore=0, mailbox=0xf6e920a0 "", 
mailboxlen=0) at chan_sip.c:5844
#4  0xf6ec3129 in handle_request (p=0x929bdf0, req=0xf6e9bb50, 
sin=0xf6e9bb40, recount=0x0, nounlock=0xf6e9b9c8) at chan_sip.c:8384
#5  0xf6ec5281 in sipsock_read (id=0x927cc50, fd=13, events=1, 
ignore=0x0) at chan_sip.c:8598
#6  0x0805378f in ast_io_wait (ioc=0x927cc10, howlong=0) at io.c:267
#7  0xf6ec89b2 in do_monitor (data=0x0) at chan_sip.c:8745
#8  0x008661d5 in start_thread () from /lib/tls/libpthread.so.0
#9  0x007c02da in clone () from /lib/tls/libc.so.6

Thanks,
JO



Matthew Boehm wrote:
>>#1  0xf6eb58c0 in build_peer (name=0x0, v=0x95fa370, realtime=0)
>>    at chan_sip.c:9255
>>#2  0xf6eb67b0 in find_peer (peer=0x0, sin=0x9642fd4, realtime=1)
>>    at chan_sip.c:1222
>>    
>>
>
>    I just updated to newest CVS and visited those line numbers above. 9255
>is "in" build_peer but find_peer is no where near 1222.
>
>    Update to newest CVS. Recompile and try again. If it crashes just send
>what you sent that last time. That was good.
>
>-Matthew
>
>_______________________________________________
>Asterisk-Users mailing list
>Asterisk-Users@lists.digium.com
>http://lists.digium.com/mailman/listinfo/asterisk-users
>To UNSUBSCRIBE or update options visit:
>   http://lists.digium.com/mailman/listinfo/asterisk-users
>  
>

-- 
Jose R. Ortiz Ubarri (CHEO), CS
System Administrator / Programmer
High Performance Computing facility - UPR
Email: Jose.Ortiz@hpcf.upr.edu|cheo@hpcf.upr.edu
Phone: 787-758-3054
Fax: 787-758-3058

Jose R. Ortiz Ubarri wrote:
> Hi:
>    I had asterisk with RealTime database working perfectly in a RH 9.0 
> machine.  I used the sip cache so I even had MWI working.  The problem 
> is that I decided to move to Fedora Core 3.  I installed the lastets 
> cvs version of asterisk and the RealTime addon from asterisk-addons.  
> I at first had the problems with the kernel and the zaptel driver but 
> all that was solved with the configuration from the Asterisk Wiki.  
> Then when I moved my configuration to the new asterisk server and 
> configured the RealTime addon it falls in a Segmentation fault.  If I 
> do not load the res_config_mysql.so (edited at modules.conf) then 
> asterisks runs without any problem.  But if I load the module from 
> boot or from the asterisk command load res_config_mysql.so then I get 
> the Segmentation fault again.
>
> I'm not sure what the problem is.  Is it a Fedora Core 3 problem, or 
> an Asterisk latest version problem?
> I don't think it is a configuration problem because I just used the 
> same configuration I had before.  The only diferences may be the OS 
> and probably the asterisk version that is only one week newer than the 
> one I was running in the old asterisk server, so I'm probably even 
> running the same version of asterisk in both machines.
>
> Any advise?  Someone else have a similar configuration working with 
> Fedora Core 3?
>
> Thanks in advance,
>
Debugging the code and as you can see in the backtrace the problem is 
that it is receiving a Null variable (name) and then making the 
comparison.  Is it an asterisk bug?  What asterisk should do if the 
variable name received is NULL? 


-- 
Jose R. Ortiz Ubarri (CHEO), CS
System Administrator / Programmer
High Performance Computing facility - UPR
Email: Jose.Ortiz@hpcf.upr.edu|cheo@hpcf.upr.edu
Phone: 787-758-3054
Fax: 787-758-3058