[OLSR-users] interaction between mobile ipv6 and olsr

Andreas Tønnesen (spam-protected)
Tue Nov 29 19:40:59 CET 2005


Hi Magesh,

Thanks for the info! I was able to reproduce this and I have
commited some fixes. It would be great if you could update
from CVS tomorrow and verify that olsrd no longer crashes :-)
Same goes for everybody else.

- Andreas

Magesh mannar wrote:
> Hi
> 
> Hopefully this trace might give you a hint.
> 
> let me explain my scenario
> 
> I have 2 nodes running olsrd (downloaded from the cvs
> last wee
> 
> n1----n2
> 
> n1 and n2 are neighbors. Whilst starting olsrd, both
> have ipv6 address assigned. The routing works fine 
> 
> Now i change the ipv6 address on the interface (eth1)
> on which the olsrd is running. olsrd on which the
> address is changed (n1) crashes. 
> 
> To my understanding, olsrd probes the interface and
> get the new ip address and  tries to schedule events.
> The it crashes with seg fault. When i do bt, it says
> program no longer exists and no stack
> I have attached my gdb trace file as well. 
> 
> 
> This problem seem to exist for a long time. Might be
> some data structure is not getting intiliased
> properly.
> 
> Probably, this is also related to the ignacio's
> problem
> 
> If you need more details,i can provide you
> 
> Magesh
>  
> 
> 
> 
> 
> Now
> 
> --- Andreas Tønnesen <(spam-protected)> wrote:
> 
> 
>>We would really like to have this issue fixed in
>>0.4.10. If the
>>core dump is too much trouble a gdb backtrace would
>>also be
>>of help.
>>Here are some easy steps:
>>
>>gdb olsrd
>>run
>>
>>Now when olsrd crashes do
>>bt
>>
>>and post the output of the entire session on the
>>list. Also the
>>configuration file would be useful. If you could use
>>the current
>>CVS version that would be even better!
>>This would be really helpfull :-) If others have
>>seen the same
>>issue, feel free to reply to this.
>>
>>- Andreas
>>
>>Andreas Tønnesen wrote:
>>
>>>Hi Ignacio,
>>>
>>>Thanks for reporting this. Here is Thomas' core
>>
>>dump recepy, could
>>
>>>you follow it and post the files?
>>>Thanks.
>>>
>>>- Andreas
>>>
>>>1. Change to the olsrd source code directory.
>>>
>>>2. Say "make uberclean; make" to force a complete
>>
>>rebuild
>>
>>>and to build olsrd.
>>>
>>>3. Do NOT say "make install". "make install" would
>>
>>strip the debug symbols
>>
>>>from the compiled executable and from the plug-ins
>>
>>in order to reduce the
>>
>>>size of the executables and the plug-ins. But we
>>
>>need the debug symbols.
>>
>>>4. Manually copy the olsrd executable and the
>>
>>plug-in .so files(if you
>>
>>>use any) to the appropriate directories.
>>
>>("/usr/sbin" and "/usr/lib" by
>>
>>>default)
>>>
>>>5. You now have an olsrd executable with debug
>>
>>information and plug-ins
>>
>>>with debug information installed.
>>>
>>>6. Say "ulimit -c unlimited". This will allow
>>
>>olsrd to dump core.
>>
>>>7. Run olsrd and make it crash.
>>>
>>>8. Create a TAR of the (1) olsrd executable, (2)
>>
>>the plug-in .so files,
>>
>>>(3) the core file, and send it to me. This will
>>
>>enable me to find out
>>
>>>where exactly the crash happens.
>>>
>>>Thanks,
>>>-Thomas
>>>
>>>Ignacio García Pérez wrote:
>>>
>>>
>>>>Hi,
>>>>
>>>>I'm using olsrd 0.4.9, and noticed that when I
>>
>>take down an interface
>>
>>>>that is part of the mesh, it segfaults most of
>>
>>the time. This is the
>>
>>>>last it shows with the debug verbosity on:
>>>>
>>>>************** TABLES ****************
>>>>Routing table:
>>>>ROUTING TABLE
>>>>DESTINATION     NEXT HOP
>>>>Old table:
>>>>ROUTING TABLE
>>>>DESTINATION     NEXT HOP
>>>>**************************************
>>>>Updating kernel routes...
>>>>Updating kernel HNA routes...
>>>>HNA table:
>>>>ROUTING TABLE
>>>>DESTINATION     NEXT HOP
>>>>mid set: 13:05:59.709971
>>>>10.50.2.1:  10.50.2.40
>>>>
>>>>DUP TABLE:
>>>>
>>>>
>>>>--- 13:05:59.70
>>
>>-------------------------------------------------
>>HNA SET
>>
>>>>IP net          netmask         GW IP
>>>>
>>>>--- 13:05:59.70
>>
>>----------------------------------------------------
>>
>>
>>>>LINKS
>>>>
>>>>IP address       hyst   LQ     lost   total  NLQ 
>>
>>  ETX
>>
>>>>10.50.2.40       0.000  0.000  0      0     
>>
>>0.000  0.00
>>
>>>>--- 13:05:59.70
>>
>>------------------------------------------------ 
>>
>>>>NEIGHBORS
>>>>
>>>>IP address       LQ     NLQ    SYM   MPR   MPRS 
>>
>>will
>>
>>>>Segmentation fault
>>>>
>>>>
>>>>
>>>>Nacho.
>>>>
>>>>
>>>>
>>>>_______________________________________________
>>>>olsr-users mailing list
>>>>(spam-protected)
>>>>https://www.olsr.org/mailman/listinfo/olsr-users
>>>
>>>
>>-- 
>>Andreas Tønnesen
>>http://www.olsr.org
>>http://www.tønnesen.org
>>
>>_______________________________________________
>>olsr-users mailing list
>>(spam-protected)
>>https://www.olsr.org/mailman/listinfo/olsr-users
>>
> 
> 
> 
> 	
> 		
> __________________________________ 
> Yahoo! Mail - PC Magazine Editors' Choice 2005 
> http://mail.yahoo.com

-- 
Andreas Tønnesen
http://www.olsr.org
http://www.tønnesen.org




More information about the Olsr-users mailing list