From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marcel Holtmann Subject: Re: [PATCH 00/39] merge request for WiMAX kernel stack and i2400m driver v2 Date: Thu, 27 Nov 2008 11:47:17 +0100 Message-ID: References: Mime-Version: 1.0 (Apple Message framework v929.2) Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, wimax@linuxwimax.org To: Inaky Perez-Gonzalez Return-path: Received: from senator.holtmann.net ([87.106.208.187]:56813 "EHLO mail.holtmann.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755521AbYK0Kr0 (ORCPT ); Thu, 27 Nov 2008 05:47:26 -0500 In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: Hi Inaky, > This code is broken up in a WiMAX subsystem and a driver for the Intel > 2400m Wireless WiMAX Link. To be able to fully utilize WiMAX services, > you need extra pieces of device-specific user space software; for now, > there is a C level API (the Common API) to use/control it, but a high > level API over DBUS that should work for any device is in the > works. See details further below. this is not a big deal, but a more general question on how the layout of the kernel source code tree. So the whole WiMAX subsystem should be in net/wimax/, but for the drivers I think using drivers/net/wimax/ is wrong. None of these drivers should ever register an Ethernet or network device directly. They register themselves with the WiMAX subystem. So I would propose to use drivers/wimax/ here. We do the same with Bluetooth. All the drivers are under drivers/bluetooth/, because they don't register any kind of network device. Jeff, what is your take on this. Shouldn't be drivers/net/ only contain drivers that actually use struct net_device somehow? Regards Marcel