[ntp:hackers] restrict and IPv6 issues

Michael Tatarinov kukabu at gmail.com
Tue May 19 06:35:37 UTC 2009


Hello

I think a good idea to abandon "restrict -4 default" and "restrict -6
default" and use a single parser for IPv4/IPv6 addresses.

Dave, I use a tunnel broker (http://www.tunnelbroker.net/) to test
IPv6, also Windows native supported Teredo (IPv6 over UDP)

2009/5/19, Dave Hart <davehart at gmail.com>:
> FYI, please feel free to join the discussion.  It's archived at
> http://lists.ntp.org/pipermail/hackers/
>
> Cheers,
> Dave Hart
>
>
> ---------- Forwarded message ----------
> From: Dave Hart <davehart at gmail.com>
> Date: Tue, May 19, 2009 at 5:52 AM
> Subject: restrict and IPv6 issues
> To: "hackers at lists.ntp.org" <hackers at lists.ntp.org>
>
>
> http://bugs.ntp.org/1192 was just filed pointing out that ntpd
> 4.2.5p177 throws a syntax error given a ntp.conf directive starting
> with:
>
> restrict -6 default
>
> I assume that "restrict -6 default" used to work, based on the bug
> report.  In glancing at the code, I see another problem that I
> consider a bug, but I'm interested in hearing from anyone who
> disagrees.  ntpd 4.2.5 maintains two restriction lists, one for IPv4,
> and one for IPv6.  Currently, "restrict default ..." applies to only
> one protocol or the other, IPv4 by default, IPv6 if ntpd was invoked
> with the -6 switch.  It seems to me "restrict default" should apply to
> both protocols.  Once bug 1192 is fixed and "restrict -4 default" and
> "restrict -6 default" are again supported, they should of course apply
> to only the given protocol.
>
> As an aside, my experience with ntpd on IPv6 is extremely limited,
> frustratingly, as over two years after Martin got ntpd IPv6 support
> working on Windows, that code is still in limbo.  I don't run any
> Unix-based machines to test ntpd IPv6 support on myself, though a kind
> NTP user has given me a shell on their Linux box and set me up with
> sudo to be able to run ntpd, however they do not use IPv6.  I'm
> comfortable mucking around in ntp_scanner.c and ntp_parser.y and
> intend to fix 1192 and the bug I identified, but it sure as heck isn't
> to scratch a current itch.  Hopefully it won't be too many more years
> before I can test ntpd IPv6 stuff on my Windows machines.
>
> Cheers,
> Dave Hart
>


More information about the hackers mailing list