From mboxrd@z Thu Jan 1 00:00:00 1970 From: Joseph Gasparakis Subject: Re: [net-next v4 07/16] i40e: Rx checksum offload for VXLAN Date: Fri, 3 Jan 2014 11:07:39 -0800 (PST) Message-ID: References: <1388200940-9113-1-git-send-email-jeffrey.t.kirsher@intel.com> <1388200940-9113-8-git-send-email-jeffrey.t.kirsher@intel.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Cc: Joseph Gasparakis , "Kirsher, Jeffrey T" , David Miller , "netdev@vger.kernel.org" , "gospo@redhat.com" , "sassmann@redhat.com" , "Brandeburg, Jesse" To: Or Gerlitz Return-path: Received: from mga03.intel.com ([143.182.124.21]:36251 "EHLO mga03.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751158AbaACStd (ORCPT ); Fri, 3 Jan 2014 13:49:33 -0500 In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On Thu, 2 Jan 2014, Or Gerlitz wrote: > On Thu, Jan 2, 2014 at 10:57 PM, Joseph Gasparakis > wrote: > > > > > > The comment says that the driver has to compute the udp checksum > > > **if** the packet has an outer udp checksum, > > > but the driver call to issue checksumming is done unconditionally > > > > > > This code runs when the right bits are set in the rx hardware descriptor > > field (rx_ptype and rx_status), which implies the hardware has detected > > ipv4 tunneling and there is non zero UDP csum. > > > Got it -- thanks for explaining, also for the pointer on where > skb->encapsulation is set > Any time Or. And thanks for your reviews.