From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail01a.mail.t-online.hu ([84.2.40.6]:55055 "EHLO mail01a.mail.t-online.hu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753543AbZKAV7J (ORCPT ); Sun, 1 Nov 2009 16:59:09 -0500 Message-ID: <4AEE04AA.1040308@freemail.hu> Date: Sun, 01 Nov 2009 22:59:06 +0100 From: =?ISO-8859-1?Q?N=E9meth_M=E1rton?= MIME-Version: 1.0 To: Jean-Francois Moine CC: Hans de Goede , V4L Mailing List , Thomas Kaiser , Theodore Kilgore , Kyle Guinn Subject: Re: [PATCH 00/21] gspca pac7302/pac7311: separate the two drivers References: <4AECC486.7080404@freemail.hu> <20091101095259.67122ef1@tele> In-Reply-To: <20091101095259.67122ef1@tele> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: linux-media-owner@vger.kernel.org List-ID: Jean-Francois Moine wrote: > On Sun, 01 Nov 2009 00:13:10 +0100 > Németh Márton wrote: > >> the following patchset refactores the Pixart PAC7311 subdriver. The >> current situation is that the code contains a lot of decisions >> like this: >> >> if (sd->sensor == SENSOR_PAC7302) { >> ... do this ... >> } else { >> ... do something else ... >> } >> >> The sensor type is determined using the USB Vendor ID and Product >> ID which means that the decisions shown are not really necessary. >> >> The goal of the patchset is to have a PAC7302 and a PAC7311 subdriver >> which have the benefit that there is no decision necessary on sensor >> type at runtime. The common functions can be extracted later but this >> would be a different patchset. > > Splitting the pac7311 subdriver is a good idea, but I don't like your > patchset: a lot of changes (function prefixes) are nullified by the > last patch. I'd better like only one change for the pac7302 creation > and a second one for the interface change of pac_find_sof() in > pac_common.h (BTW, this file could now be compiled separately). Hello Jef, thank you for the feedback, I'll try to send a patch set wich contains bigger steps. I hope the separation will be not a too big step and won't make it too difficult to bisect any possible problem I might introduce with this change. But hope for the best and imagine the easy way when no regression was introduced. I am also thinking about finding the common functions which can be compiled separately either in a helper module or to gspca_main maybe. But first I focus on the pac7302/pac7311 separation. Márton Németh