From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Yonan Subject: Question on bulk dequeue support in virtual drivers Date: Thu, 20 Nov 2014 05:39:44 -0700 Message-ID: <546DE110.8060607@openvpn.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit To: Linux Netdev List Return-path: Received: from mail.yonan.net ([54.244.116.145]:58441 "EHLO mail.yonan.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751894AbaKTMjp (ORCPT ); Thu, 20 Nov 2014 07:39:45 -0500 Received: from Moab.local (c-67-166-32-18.hsd1.co.comcast.net [67.166.32.18]) (using TLSv1 with cipher ECDHE-RSA-AES128-SHA (128/128 bits)) (No client certificate requested) by mail.yonan.net (Postfix) with ESMTPSA id 211EAC0E5D for ; Thu, 20 Nov 2014 05:39:45 -0700 (MST) Sender: netdev-owner@vger.kernel.org List-ID: Consider a tunneling driver that receives packets in ndo_start_xmit, encapsulates them in UDP, and forwards via ip_local_out. The ndo_start_xmit implementation can implement bulking by looking at skb->xmit_more. But then how to efficiently transmit the resulting bulked list of encapsulated packets via UDP, so that the packets both enter and leave the driver in bulked form? Is there an ip_local_out alternative for bulked skb lists? James