| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <DUB107-W47D427AB8714864D88CD46EDBA0@phx.gbl> Date: Tue, 7 May 2013 20:15:36 +0200 From: Lukas Tribus <luky-37@...mail.com> To: Michael Tokarev <mjt@....msk.ru>, netdev <netdev@...r.kernel.org> Subject: RE: source IP address for ARP requests > eth0 10.0.0.1/32 > eth1 192.168.1.1/24 > route add 10.0.0.2 dev eth1 src 10.0.0.1 > > namely, so that just one ip address is available > on an "unrelated" subnet, with "our" address taken > from another interface. > > With this config, ARPs for 10.0.0.2 are sent to eth1 > with source address of 192.168.1.1, even if "src" > is configured for the route in question: > > 9:28:46.220680 ARP, Request who-has 10.0.0.2 tell > 192.168.1.1, length 28 > > Is that right, or should the host send ARP requests > with src=10.0.0.1? I suspect you cannot change the behavior of layer 2 discovery protocols by configuring a certain layer 3 parameter (a route + "src"). Also from an ARP perspective, this is imho correct behavior, because 10.0.0.2 should learn "sender hardware address" and "sender protocol address" from the ARP _request_ and add them to its own ARP table - without an additional ARP request. 10.0.0.2 is connected to eth1, so only mac and ip address of eth1 counts in this case (= 192.168.1.1). Regards, Lukas -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@...r.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists