From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jiri Benc Subject: Re: [PATCH net-next v2 4/7] vxlan: improve vxlan route lookup checks. Date: Wed, 9 Nov 2016 17:41:38 +0100 Message-ID: <20161109174138.410b4385@griffin> References: <1478371557-71888-1-git-send-email-pshelar@ovn.org> <1478371557-71888-5-git-send-email-pshelar@ovn.org> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org To: Pravin B Shelar Return-path: Received: from mx1.redhat.com ([209.132.183.28]:43618 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754107AbcKIQlk (ORCPT ); Wed, 9 Nov 2016 11:41:40 -0500 In-Reply-To: <1478371557-71888-5-git-send-email-pshelar@ovn.org> Sender: netdev-owner@vger.kernel.org List-ID: On Sat, 5 Nov 2016 11:45:54 -0700, Pravin B Shelar wrote: > Move route sanity check to respective vxlan[4/6]_get_route functions. > This allows us to perform all sanity checks before caching the dst so > that we can avoid these checks on subsequent packets. > This give move accurate metadata information for packet from > fill_metadata_dst(). The description is misleading, it applies only to one vxlan lwt use case (openvswitch). For other use cases, the patch has no effect. I found the current handling of route lookup results irritating, too. The reason I did not change this while doing vxlan cleanup some time ago was that I assumed we should not increase dev stats from vxlan_fill_metadata_dst. Isn't that so? Jiri