From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Barnhill <0xeffeff@gmail.com> Subject: Re: [PATCH net v5] net/ipv6: Add anycast addresses to a global hashtable Date: Wed, 31 Oct 2018 20:02:30 -0400 Message-ID: References: <20181030.113147.173837020180526004.davem@davemloft.net> <75a6a4bb-a8f3-4996-a157-825a5834ae52@gmail.com> <20181030.161916.2155476722804506340.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Cc: David Ahern , netdev@vger.kernel.org, Alexey Kuznetsov , yoshfuji@linux-ipv6.org To: davem@davemloft.net Return-path: Received: from mail-yb1-f196.google.com ([209.85.219.196]:42983 "EHLO mail-yb1-f196.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726227AbeKAJDZ (ORCPT ); Thu, 1 Nov 2018 05:03:25 -0400 Received: by mail-yb1-f196.google.com with SMTP id o204-v6so7419322yba.9 for ; Wed, 31 Oct 2018 17:02:58 -0700 (PDT) In-Reply-To: <20181030.161916.2155476722804506340.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: I'll follow this email with a new patch using ifacaddr6 instead of creating a new struct. I ended up using fib6_nh.nh_dev to get the net, instead of adding a back pointer to idev. It seems that idev was recently removed in lieu of this, so if this is incorrect, please let me know. Hopefully, I got the locking correct. Thanks, Jeff On Tue, Oct 30, 2018 at 7:19 PM David Miller wrote: > > From: David Ahern > Date: Tue, 30 Oct 2018 16:06:46 -0600 > > > or make the table per namespace. > > This will increase namespace create/destroy cost, so I'd rather not > for something like this.