Unidentified subject!

	<yoshfuji@linux-ipv6.org>  Fri Apr  5 21:13:27 2002
X-Envelope-Sender: yoshfuji@linux-ipv6.org
Received: (qmail 1869 invoked from network); 6 Apr 2002 03:13:25 -0000
Received: from yue.hongo.wide.ad.jp (root@203.178.139.94)
  by murphy.debian.org with SMTP; 6 Apr 2002 03:13:25 -0000
Received: from localhost (yoshfuji@localhost [127.0.0.1])
	by yue.hongo.wide.ad.jp (8.9.3+3.2W/8.9.3/Debian 8.9.3-21) with ESMTP id MAA06554;
	Sat, 6 Apr 2002 12:13:07 +0900
To: djm@mindrot.org
Cc: matthew@debian.org, wichert@wiggy.net, debian-ipv6@lists.debian.org,
        debian-ssh@lists.debian.org
Subject: Re: IPv6 support
In-Reply-To: <[🔎] Pine.LNX.4.44.0204061057390.1260-100000@mothra.mindrot.org>
References: <20020302134214W.yoshfuji@wide.ad.jp>
	<[🔎] Pine.LNX.4.44.0204061057390.1260-100000@mothra.mindrot.org>
X-Mailer: Mew version 1.94.2 on Emacs 20.7 / Mule 4.1 (AOI)
Sender: "Hideaki YOSHIFUJI" <yoshfuji@cerberus.hongo.wide.ad.jp>
X-URL: http://www.yoshifuji.org/%7Ehideaki/
X-Fingerprint: 90 22 65 EB 1E CF 3A D1 0B DF 80 D8 48 07 F8 94 E0 62 0E EA
X-PGP-Key-URL: http://www.yoshifuji.org/%7Ehideaki/hideaki@yoshifuji.org.asc
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-Id: <20020406121307E.yoshfuji@linux-ipv6.org>
Date: Sat, 06 Apr 2002 12:13:07 +0900
From: YOSHIFUJI Hideaki / =?iso-2022-jp?B?GyRCNUhGIzFRTEAbKEI=?= 
	<yoshfuji@linux-ipv6.org>
X-Dispatcher: imput version 991025(IM133)
Lines: 41
X-Spam-Status: No, hits=-2.0 required=4.7 tests=IN_REP_TO version=2.01

In article <[🔎] Pine.LNX.4.44.0204061057390.1260-100000@mothra.mindrot.org> (at Sat, 6 Apr 2002 11:02:29 +1000 (EST)), Damien Miller <djm@mindrot.org> says:

> On Sat, 2 Mar 2002, YOSHIFUJI Hideaki / [iso-2022-jp] 吉藤英明 wrote:
> 
> > Please apply this fix (against openssh-3.0.2p1) to suppress an 
> > "error" message which have people confused because it is a warning 
> > in fact.
> 
> I would prefer to make the test still more specific - could it check
> for EINVAL returns from bind() only?

It is ok to ignore the "error" if bind(2) for ipv4 socket returns
EADDRINUSE error when you have ipv6 socket bound. 
Similarly, it is ok to ignore the "error" if bind(2) for ipv6 socket 
returns EADDRINUSE error when you have ipv4 socket bound.

The loop trys only for ipv4 and ipv6 socket.  So, this compiles to:
 It is ok to ignore the "error" if bind(2) for a socket returns
 EADDRINUSE error then you have another socket bound.

We may need to check IPV6_V6ONLY socket option support.
If IPV6_V6ONLY is defined, set IPV6_V6ONLY option to (int)1
for ipv6 socket to have ipv4 bound.  And, above "error" is 
a real error if IPV6_V6ONLY is defined and supported (setsockopt() 
succeeded).


summary:
 0. let v4inv6 0 outside the loop.
 1. set IPV6_V6ONLY for ipv6 socket before bind(2) if it is defined.
    if IPV6_V6ONLY is defined and setsockopt() failed, warn it.
    if IPV6_V6ONLY is defined and setsockopt() succeeded, let v4inv6 1.

 2. if bind(2) for an socket (either for ipv4 or ipv6) failed when you 
    already have socket (it is either ipv6 or ipv4, respectively because the 
    loop trys only for ipv4 and/or ipv6) and the reason is EADDRINUSE and 
    v4inv6 is 0, it is ok to ignore this "error."

-- 
Hideaki YOSHIFUJI @ USAGI Project <yoshfuji@linux-ipv6.org>
GPG FP: 9022 65EB 1ECF 3AD1 0BDF  80D8 4807 F894 E062 0EEA


-- 
To UNSUBSCRIBE, email to debian-ssh-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org