From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752960AbaIBD0h (ORCPT ); Mon, 1 Sep 2014 23:26:37 -0400 Received: from shards.monkeyblade.net ([149.20.54.216]:46770 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751832AbaIBD0f (ORCPT ); Mon, 1 Sep 2014 23:26:35 -0400 Date: Mon, 01 Sep 2014 20:26:32 -0700 (PDT) Message-Id: <20140901.202632.1784070145688270115.davem@davemloft.net> To: antoine.tenart@free-electrons.com Cc: sebastian.hesselbarth@gmail.com, thomas.petazzoni@free-electrons.com, alexandre.belloni@free-electrons.com, zmxu@marvell.com, jszhang@marvell.com, netdev@vger.kernel.org, linux-arm-kernel@lists.infradead.org, devicetree@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH 1/5] net: mvberlin_eth: add an Ethernet driver for Marvell Berlin From: David Miller In-Reply-To: <1409320263-10295-2-git-send-email-antoine.tenart@free-electrons.com> References: <1409320263-10295-1-git-send-email-antoine.tenart@free-electrons.com> <1409320263-10295-2-git-send-email-antoine.tenart@free-electrons.com> X-Mailer: Mew version 6.5 on Emacs 24.1 / Mule 6.0 (HANACHIRUSATO) Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Greylist: Sender succeeded SMTP AUTH, not delayed by milter-greylist-4.5.7 (shards.monkeyblade.net [149.20.54.216]); Mon, 01 Sep 2014 20:26:35 -0700 (PDT) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Antoine Tenart Date: Fri, 29 Aug 2014 15:50:59 +0200 > +struct rx_desc { > + u16 buf_size; /* Buffer size */ > + u16 byte_cnt; /* Descriptor buffer byte count */ ... > + u16 byte_cnt; /* buffer byte count */ > + u16 l4i_chk; /* CPU provided TCP checksum */ I'm really suspcious that you're ordering these two members differently than that used by the mv643xx_eth driver. Looking at a diff of the two drivers, most of the transmit and recieve paths are largely the same. WRT. register differences, adding indirection to handle that is a no-brainer because the cost of the register access itself will absolutely dwarf whatever it costs to do an indirect call or whatever to implement the register access. I really think you should look into sharing code more seriously.