Error in dig 9.6.1-P2
marka at isc.org
Wed Dec 15 21:32:31 UTC 2010
In message <4D090672.8020004 at isc.org>, Michael Graff writes:
> On 2010-12-15 12:02 PM, Sara Monteiro wrote:
> > Hi,
> > I have this error some times in my VM in CentOS:
> > # dig org. ANY
> > ;; Truncated, retrying in TCP mode.
> > socket.c:2486: REQUIRE((((sock) != ((void *)0)) && (((const isc__magic_t
> > *)(sock))->magic == ((('I') << 24 | ('O') << 16 | ('i') << 8 |
> > ('o')))))) failed.
> > Aborted
> > but if I do the same query with +dnssec it works, does any one knows why
> > this happens?
> It's clearly a bug. Can you send it in to bind9-bugs at isc.org?
> the REQUIRE() above is likely indicating that we (as programmers) have
> passed in a NULL pointer somewhere. If you have a core dump, that would
> be useful too, ideally a stack trace.
It's also almost certainly fixed in more recent code. BIND 9.6.1-P2
is old. BIND 9.6.3b1 is about to be released and it contains a
number of fixes to bugs with symptoms like this if you want to stay
with BIND 9.6.x. BIND 9.7.2-P3 already has the fixes.
As to why it doesn't fail with +dnssec is that the answer fits in
the EDNS packet so there is no fallback to TCP.
> - --Michael
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.11 (Darwin)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
> -----END PGP SIGNATURE-----
> bind-workers mailing list
> bind-workers at lists.isc.org
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742 INTERNET: marka at isc.org
More information about the bind-workers