httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Colm MacCarthaigh <c...@stdlib.net>
Subject Re: 2.1.7 Available for Testing & Voting
Date Thu, 25 Aug 2005 14:17:37 GMT
On Thu, Aug 25, 2005 at 10:11:01AM -0400, Jeff Trawick wrote:
> first entry:
> AF_INET; 0.0.0.0:8080
> second (last) entry:
> AF_INET6; [::]:8080

Yuck, We'll have to re-order the list so, or change the listen.c logic.

> But if it is the other way around:
> 
> Get AF_INET6 socket first and set IPV6_V6ONLY to 0 (--enable-v4-mapped
> is the default on non *bsd).  Then get AF_INET socket and try to bind.
>  Why is that guaranteed to work since we have overlap?

See lines 391 -> 405 of server/listen.x, the AF_INET socket will be
skipped in such a circumstance.

> Seems to me that we should scrap the --enable-v4-mapped stuff, default
> to getting two listening sockets everywhere (if IPv6 enabled, of
> course), and (future) have a Listen option to get a single socket to
> handle mapped addresses. 

We can't default to having two listening sockets everywhere, there are
well-deployed operating systems which simply don't support IPV6_V6ONLY.

> With that Listen option we'd just pass AF_INET6 instead of AF_UNSPEC
> on the resolver call since the admin who specified that option knows
> what they're doing.

Specified what option?

> (Why bother with mapped addresses? 

Because they are the only way to accept IPv6 at all on many systems
(like most Linux installs, for example).

-- 
Colm MacCárthaigh                        Public Key: colm+pgp@stdlib.net

Mime
View raw message