From mboxrd@z Thu Jan 1 00:00:00 1970 From: roopa Subject: Re: [PATCH net v2] switchdev: don't abort hardware ipv4 fib offload on failure to program fib entry in hardware Date: Mon, 18 May 2015 22:57:16 -0700 Message-ID: <555AD0BC.7000608@cumulusnetworks.com> References: <1431906125-13808-1-git-send-email-roopa@cumulusnetworks.com> <20150518.161916.2132217836491222672.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: sfeldma@gmail.com, john.fastabend@gmail.com, jiri@resnulli.us, netdev@vger.kernel.org To: David Miller Return-path: Received: from mail-pa0-f44.google.com ([209.85.220.44]:34501 "EHLO mail-pa0-f44.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751218AbbESF5S (ORCPT ); Tue, 19 May 2015 01:57:18 -0400 Received: by pabru16 with SMTP id ru16so8981122pab.1 for ; Mon, 18 May 2015 22:57:18 -0700 (PDT) In-Reply-To: <20150518.161916.2132217836491222672.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: On 5/18/15, 1:19 PM, David Miller wrote: > From: Roopa Prabhu > Date: Sun, 17 May 2015 16:42:05 -0700 > >> On most systems where you can offload routes to hardware, >> doing routing in software is not an option (the cpu limitations >> make routing impossible in software). > You absolutely do not get to determine this policy, none of us > do. > > What matters is that by default the damn switch device being there > is %100 transparent to the user. > > And the way to achieve that default is to do software routes as > a fallback. > > I am not going to entertain changes of this nature which fail > route loading by default just because we've exceeded a device's > HW capacity to offload. > > I thought I was _really_ clear about this at netdev 0.1 sorry if i missed you emphasize this. Current fib hw offload code aborts hw offload on the first hw offload failure and moves all routes to software and before 4.1 goes out I was trying to revisit this because the existing defaults will not work for most devices. To that effect, i started with an RFC patch to get some current thoughts and hence this patch. I will follow up on the other options mentioned in the comment.