From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ido Schimmel Subject: Re: [PATCH net-next 3/4] ipv6: Use hash-threshold instead of modulo-N Date: Wed, 10 Jan 2018 14:02:50 +0200 Message-ID: <20180110120250.GB12249@splinter> References: <20180109144028.30133-1-idosch@mellanox.com> <20180109144028.30133-4-idosch@mellanox.com> <397562a6-35ec-c5e4-b819-9931c1e5d86f@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Ido Schimmel , netdev@vger.kernel.org, davem@davemloft.net, roopa@cumulusnetworks.com, nikolay@cumulusnetworks.com, pch@ordbogen.com, jkbs@redhat.com, yoshfuji@linux-ipv6.org, mlxsw@mellanox.com To: David Ahern Return-path: Received: from out3-smtp.messagingengine.com ([66.111.4.27]:32903 "EHLO out3-smtp.messagingengine.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752766AbeAJMCy (ORCPT ); Wed, 10 Jan 2018 07:02:54 -0500 Content-Disposition: inline In-Reply-To: <397562a6-35ec-c5e4-b819-9931c1e5d86f@gmail.com> Sender: netdev-owner@vger.kernel.org List-ID: On Tue, Jan 09, 2018 at 08:54:00PM -0700, David Ahern wrote: > Did you consider adding the net->ipv4.sysctl_fib_multipath_use_neigh > check as well? ie., if set, verify neighbor is alive before picking that hop No, but I'll add it to my TODO list. It's what effectively happens in mlxsw. If the neighbour isn't NUD_VALID, then we don't have a MAC we can use in the adjacency table and the nexthop isn't considered during multipath selection.