On Mon, 21 Dec 2015, Donald Sharp wrote:
I applied the patch and ran my smoke test here is the new core that I am
getting:
[New LWP 5727]
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/powerpc-linux-gnu/libthread_db.so.1".
Core was generated by `/usr/lib/quagga/ospfd --daemon -A 127.0.0.1'.
Program terminated with signal 11, Segmentation fault.
#0 0x0ffc0530 in ?? () from /usr/lib/libospf.so.0
#0 0x0ffc0530 in ?? () from /usr/lib/libospf.so.0
#1 0x0ff89270 in ospf_nsm_event (thread=<optimized out>) at ospf_nsm.c:791
#2 0x0ff11820 in thread_call (thread=thread@entry=0xbfc4c308) at
thread.c:1266
#3 0x0ff1194c in funcname_thread_execute (m=<optimized out>,
func=0xff891a0 <ospf_nsm_event>, arg=0x1007b980, val=val@entry=11,
funcname=funcname@entry=0xffbdcc6 "ospf_nsm_event",
schedfrom=schedfrom@entry=0xffbe63b "ospf_interface.c",
fromln=fromln@entry=310)
at thread.c:1323
#4 0x0ff84e10 in ospf_if_cleanup (oi=0x10079cc0) at ospf_interface.c:310
#5 0x0ff86c14 in ism_interface_down (oi=<optimized out>) at ospf_ism.c:394
#6 0x0ff86d08 in ospf_ism_event (thread=<optimized out>) at ospf_ism.c:616
#7 0x0ff11820 in thread_call (thread=thread@entry=0xbfc4c4d8) at
thread.c:1266
#8 0x0ff1194c in funcname_thread_execute (m=<optimized out>,
func=0xff86cb8 <ospf_ism_event>, arg=arg@entry=0x10079cc0, val=val@entry=7,
funcname=funcname@entry=0xffbe62c "ospf_ism_event",
schedfrom=schedfrom@entry=0xffbe63b "ospf_interface.c",
fromln=fromln@entry=848)
at thread.c:1323
#9 0x0ff85b94 in ospf_if_down (oi=oi@entry=0x10079cc0) at
ospf_interface.c:848
#10 0x0ff85bec in ospf_if_free (oi=oi@entry=0x10079cc0) at
ospf_interface.c:329
#11 0x0ff8280c in ospf_interface_address_delete (command=<optimized out>,
zclient=<optimized out>, length=<optimized out>, vrf_id=<optimized out>) at
ospf_zebra.c:318
#12 0x0ff2535c in zclient_read (thread=<optimized out>) at zclient.c:1131
Hmm, as per IRC before you pasted this, my previous patch likely is
papering over deeper problems in the underlying code.
Presumably something is tearing things downs in a way that makes deep
assumptions that are fragile. I'm guessing in nbr_self handling. Not the
first time we've had issues here, inc. recently. Probably good to find a
way to make this code a lot more robust.
regards,
--
Paul Jakma [email protected] @pjakma Key ID: 64A2FF6A
Fortune:
Conway's Law:
In any organization there will always be one person who knows
what is going on.
This person must be fired.
_______________________________________________
Quagga-dev mailing list
[email protected]
https://lists.quagga.net/mailman/listinfo/quagga-dev
