thread crashed on linux kernel 2.6.16.13

[JINMEI Tatuya / 神明達哉]

At Wed, 16 Jul 2008 00:15:30 +0200,
Tom Mueller <t.mueller at strato-rz.de> wrote:

> bind-9.4.3b2 crashed in a thread after running for approx. 5 hrs. under
> pretty high load.
> 
> 15-Jul-2008 17:31:58.692 general: resolver.c:5494: REQUIRE((((query) !=
> ((void *)0)) && (((const isc__magic_t *)(query))->magic == ((('Q') << 24
> | ('!') << 16 | ('!') << 8 | ('!')))))) failed
> 
> Core was generated by `/opt/bind-9.4.3b2-test/sbin/named -c
> /etc/named/named.conf -n 2 -u named'.
> Program terminated with signal 6, Aborted.
> #0  0xffffe410 in __kernel_vsyscall ()
> 
> I hope this helps to fix such issues.

Thanks for reporting.  We've also received a similar crash report, but
haven't identified the cause of that (I also suspect it's somehow
related to your other report, but cannot come up with an idea about
how that could specifically happen).

Can you show the result of the following?
(gdb) thread apply all bt full
(identify the thread and stack frame of resolver.c:5494, move there,
and then)
(gdb) p *event
(gdb) p *query

---
JINMEI, Tatuya
Internet Systems Consortium, Inc.