[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
(usagi-users 03696) Re: glibc getaddrinfo can resolve addresses of different hosts in case of search domains are used in /etc/resolv.conf - bug or feature?
- To: Peter Bieringer <pb@xxxxxxxxxxxx>
- Subject: (usagi-users 03696) Re: glibc getaddrinfo can resolve addresses of different hosts in case of search domains are used in /etc/resolv.conf - bug or feature?
- From: Hajimu UMEMOTO <ume@xxxxxxxxxxxx>
- Date: Tue, 22 Aug 2006 23:35:40 +0900
- Cc: usagi-users@xxxxxxxxxxxxxx, "users@xxxxxxxx" <users@xxxxxxxx>
- In-reply-to: <44EAF1DF.6090400@bieringer.de>
- Organization: Internet Mutual Aid Society, YOKOHAMA
- References: <44EAF1DF.6090400@bieringer.de>
- Reply-to: usagi-users@xxxxxxxxxxxxxx
- Resent-date: Tue, 22 Aug 2006 23:39:10 +0900
- Resent-from: sekiya@xxxxxxxxxxxxxx
- Resent-message-id: <200608222339.FMLAAB10541.usagi-users@linux-ipv6.org>
- Resent-to: usagi-users@xxxxxxxxxxxxxx (moderated)
- User-agent: xcite1.38> Wanderlust/2.14.0 (Africa) SEMI/1.14.6 (Maruoka) FLIM/1.14.8 (Shijō) APEL/10.6 Emacs/22.0.50 (i386-pc-freebsd) MULE/5.0 (SAKAKI)
Hi,
>>>>> On Tue, 22 Aug 2006 14:00:31 +0200
>>>>> Peter Bieringer <pb@xxxxxxxxxxxx> said:
pb> after some discussions with people from Red Hat I'm still not able to
pb> convince them that the behavior of getaddrinfo in glibc is buggy, if
pb> search domains in /etc/resolv.conf are specified.
pb> Currently, it can return IPv6 and IPv4 addresses of different hosts,
pb> depending what happen during AAAA lookups while appending a search
pb> domain. If successful, application gets back e.g.
pb> AAAA fec0::1 (www.redhat.com.intranet.domain.example)
pb> A 66.187.224.150 (www.redhat.com)
pb> Not good, if application prefers IPv6...it connects unexpected to the
pb> wrong host.
pb> Me was told inbetween (and a short look into the source code shows like
pb> that), that getaddrinfo uses DNS lookups more abstract and it can't be
pb> fixed in an easy manner.
pb> Last note I get was I should provide more information or a whitepaper,
pb> that current behavior is more a bug than a feature...and support/request
pb> of the community is required.
pb> Therefore my next (last) try is to inform the IPv6 community about this
pb> issue. Please read details below and perhaps vote for
pb> ( ) bug, should be fixed in
pb> [ ] newer releases
pb> [ ] current release
pb> [ ] older releases, too
pb> ( ) feature, no need to fix it
pb> ( ) ...
pb> Feel free to add yourself to bugzilla entries shown below.
pb> BTW: would be great if one can run tests on other libc implementation
pb> like dietlibc or ulibc (or even Microsoft Windows) and report, whether
pb> one acts like the same or different. I provide a special DNS zone for
pb> that, see below.
The BSDs getaddrinfo(3) which was derived from KAME doesn't have this
problem. This problem was fixed in times past.
--
Hajimu UMEMOTO @ Internet Mutual Aid Society Yokohama, Japan
ume@xxxxxxxxxxxx ume@{,jp.}FreeBSD.org
http://www.imasy.org/~ume/