From owner-freebsd-current Mon Aug 14 03:14:16 1995 Return-Path: current-owner Received: (from majordom@localhost) by freefall.FreeBSD.org (8.6.11/8.6.6) id DAA09645 for current-outgoing; Mon, 14 Aug 1995 03:14:16 -0700 Received: from haywire.DIALix.COM (haywire.DIALix.COM [192.203.228.65]) by freefall.FreeBSD.org (8.6.11/8.6.6) with ESMTP id DAA09637 for ; Mon, 14 Aug 1995 03:14:10 -0700 Received: (from news@localhost) by haywire.DIALix.COM (8.7.Beta.11/8.7.Beta.11/DIALix) id SAA16297 for freebsd-current@freebsd.org; Mon, 14 Aug 1995 18:13:59 +0800 (WST) Received: from GATEWAY by haywire.DIALix.COM with netnews for freebsd-current@freebsd.org (problems to: usenet@haywire.dialix.com) To: freebsd-current@freebsd.org Date: 14 Aug 1995 18:13:55 +0800 From: peter@haywire.dialix.com (Peter Wemm) Message-ID: <40n7l3$ft6$1@haywire.DIALix.COM> Organization: DIALix Services, Perth, Australia. References: <40mduf$fvb$1@haywire.DIALix.COM>, <199508140542.HAA11158@uriah.heep.sax.de> Subject: Re: ioctl(SIOCAIFADDR): File exists Sender: current-owner@freebsd.org Precedence: bulk j@uriah.heep.sax.de (J Wunsch) writes: >As Peter Wemm wrote: >> >> There was a "feature" in 2.0.5 that meant that the local address of a >> point-to-point link could not be within the same network as any other >> interface, with the local subnet size being decided by the netmask of >> the remote network. >This sounds like it could be it (even though it's surprising that i >can log into the machine succesfully using another address, while i've >always been unsuccesful with the same address -- the other address has >been just 4 above). Essectially, it becomes non-deterministic.. It depends on the ordering of the routes and the interface lists.. I never really figured out exactly what the variables were, I just nuked the problem.. :-) >> There is an undocumented option in the kernel, called something like >> P2P_LOCALADDR_SHARE, which works around the problem (with a slight >> bug) after I made a lot of noise about this "feature" before 2.0.5 was >> released. >What's the `slight bug'? Can we use it nevertheless? In one of the #ifdef cases, I accidently typed if (ifp->if_flags & IFF_POINTOPOINT) if (equal(addr, ifa->ifa_addr)) ^^^^^^^^ Should be ifa_dstaddr! return (ifa); It doesn't have any affect, except possibly if you have aliases on your PPP link.. in which case, connecting to a remote address could choose the wrong local address - but that's not much different to the original code, so it's marginally better, not worse. (better, because netmasks are not applied). Cheers, -Peter >-- >cheers, J"org >joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ >Never trust an operating system you don't have sources for. ;-)