This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH][BZ 21295] getaddrinfo: do not overwrite IPv6 IPs with IPv4 when using AF_UNSPEC
- From: Florian Weimer <fweimer at redhat dot com>
- To: kmeaw at kmeaw dot com
- Cc: libc-alpha at sourceware dot org
- Date: Thu, 20 Apr 2017 16:05:01 +0200
- Subject: Re: [PATCH][BZ 21295] getaddrinfo: do not overwrite IPv6 IPs with IPv4 when using AF_UNSPEC
- Authentication-results: sourceware.org; auth=none
- Authentication-results: ext-mx02.extmail.prod.ext.phx2.redhat.com; dmarc=none (p=none dis=none) header.from=redhat.com
- Authentication-results: ext-mx02.extmail.prod.ext.phx2.redhat.com; spf=pass smtp.mailfrom=fweimer at redhat dot com
- Dkim-filter: OpenDKIM Filter v2.11.0 mx1.redhat.com 0109E9C0CB
- Dmarc-filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 0109E9C0CB
- References: <20170323132130.GA25347@kmeaw.com> <f41e3156-ccc3-2104-cc7d-90347027bb06@redhat.com> <20170420134437.GA29528@kmeaw.com>
On 04/20/2017 03:44 PM, kmeaw@kmeaw.com wrote:
But Ubuntu 12.04 does not:
root@6e6d4b17326a:/# dpkg-query -s libc6 | grep Version:
Version: 2.15-0ubuntu10.17
root@6e6d4b17326a:/# strings /lib/x86_64-linux-gnu/libnss_dns.so.2|grep name4
root@6e6d4b17326a:/# strings /lib/x86_64-linux-gnu/libnss_dns.so.2|grep name
__ns_name_unpack
__ns_name_ntop
_nss_dns_gethostbyname2_r
_nss_dns_gethostbyname_r
_nss_dns_gethostbyname3_r
__dn_skipname
_nss_dns_getnetbyname_r
_nss_dns_getcanonname_r
root@6e6d4b17326a:/#
Very odd. This is due to the patch
debian/patches/all/fedora-nss_dns-gethostbyname4-disable.diff:
# Description: disable _nss_dns_gethostbyname4_r for the moment, as it
# causes problems for IPv6 users; patch from Fedora by Jakub Jelinek
# Ubuntu: https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/313218
# Upstream: https://bugzilla.redhat.com/show_bug.cgi?id=459756
So your reproducer is indeed very Ubuntu-specific at this point. (Red
Hat Enterprise Linux 6 backported the single-request resolver option
instead, which can be used to work around issues with the parallel lookup.)
We still need to fix this upstream because it's also a bug with AI_ALL
and AI_V4MAPPED (see my other message).
Thanks,
Florian