>>>>> On Tue, 26 Jun 2001 14:30:05 +0200 (CEST),
>>>>> Erik Nordmark <[EMAIL PROTECTED]> said:
>> Just to make it sure, if you mean "accepting IPv4 packets on an
>> AF_INET6 socket as IPv4-mapped IPv6 addresses" by "the model in
>> 2553", Solaris does not follow the model, AFAIK. Also, NetBSD disable
>> the model by default.
> What aspect of this do you believe is not there is Solaris?
> Solaris by default accepts IPv4 packets on an AF_INET6 socket bound to
> in6addr_any, so there must be some sutble detail that you're referring to.
Umm, then I may have misunderstood a former discussion about itojun's
very 1st bindtest questionnaire (around August 2000). Perhaps I
misunderstood the fact that Solaris supported separate port space for
IPv4 and IPv6. Sorry, I should've been more careful when talking
about others' implementation.
By the way, I'm still collecting the differences among implementations
(see the attached message, and the latest version contains a test to
see which type of socket accepts which type of pakcets.) Could
someone try the test on Solaris (and of course other implementations
that has not been tested), and send the result to me, please? Then
I'll never misunderstand such fundamental point.
JINMEI, Tatuya
Communication Platform Lab.
Corporate R&D Center, Toshiba Corp.
[EMAIL PROTECTED]
Hello, (sorry for cross-posting).
I'm now looking at how each IPv6 implementation is friendly with
bind9, especially on the bind(2) system call ordering issues.
As the first step, I'd like to collect differences among
implementations using some test scripts. So, if you have time, could
you spend some time to do the test on your system(s)?
The test tool is freely available at the following URL:
http://orange.kame.net/dev/cvsweb.cgi/kame/kame/kame/bindtest/
and the results so far at
http://www.kame.net/newsletter/20010504/
Here is a step by step instruction of the test:
- compile bindtest (see Makefile)
- run test.sh like this:
% sh test.sh myplatform.txt
- send myplatform.txt to me, with exact indication of OS name/version
number/build date/whatever
I hear that some systems have already supported a new IPv6 socket
option "IPV6_V6ONLY". I'm particularly interested in the results on
such systems (the latest test.sh contains some tests with the option).
We'll basically put the results on the URL above, so that everyone can
refer to it. If you can do tests but do not want to make the result
open, please explicitly say so when sending the result.
Thanks in advance,
JINMEI, Tatuya
Communication Platform Lab.
Corporate R&D Center, Toshiba Corp.
[EMAIL PROTECTED]
