> after I have upgraded my ss10 (2x SM71) userland from a Juneish netbs=
d-3
> installation to NetBSD 3.1RC4, dig(1) gives me
>
> [hauke@pizza] ~ > ksh
> $ dig pizza.causeuse.org. @193.155.17.9 1>/dev/null ; echo $?
> Terminated
> 143
> $

Please try out this patch:

Index: usr.sbin/bind/include/config.h
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
RCS file: /u/nb/src/usr.sbin/bind/include/config.h,v
retrieving revision 1.3.2.1
diff -u -r1.3.2.1 config.h
--- usr.sbin/bind/include/config.h      13 Jul 2006 22:02:31 -0000     =
 1.3.2.1
+++ usr.sbin/bind/include/config.h      22 Oct 2006 15:59:32 -0000
@@ -35,8 +35,15 @@
 /* define if your system needs pthread_init() before using pthreads */=

 /* #undef NEED_PTHREAD_INIT */
 =

+/*
+ * Apparently, the combination of "no threads" but HAVE_SIGWAIT
+ * has not been well tested.  Applications end up killing themselves
+ * instead of exiting gracefully.
+ */
+#ifdef ISC_PLATFORM_USETHREADS
 /* define if your system has sigwait() */
 #define HAVE_SIGWAIT 1
+#endif /* ISC_PLATFORM_USETHREADS */
 =

 /* define if sigwait() is the UnixWare flavor */
 /* #undef HAVE_UNIXWARE_SIGWAIT */

I've tested this on an ss10 (2x SM51) running a 3.0 kernel, but
with BIND and company built from netbsd-3 sources with the above
patch, and I now get

$ dig pizza.causeuse.org. @193.155.17.9 1>/dev/null ; echo $?
0
$ =


Regards,

- H=E5vard