From mboxrd@z Thu Jan 1 00:00:00 1970 From: Greg KH Subject: Re: [PATCH 3/3] drivers/misc: introduce face detection module driver(fdif) Date: Sat, 26 Nov 2011 07:04:18 -0800 Message-ID: <20111126150418.GA11680@kroah.com> References: <1322281904-14526-1-git-send-email-tom.leiming@gmail.com> <1322281904-14526-4-git-send-email-tom.leiming@gmail.com> <20111126111247.43cb5d35@lxorguk.ukuu.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from out2.smtp.messagingengine.com ([66.111.4.26]:51786 "EHLO out2.smtp.messagingengine.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752620Ab1KZPTZ (ORCPT ); Sat, 26 Nov 2011 10:19:25 -0500 Received: from compute5.internal (compute5.nyi.mail.srv.osa [10.202.2.45]) by gateway1.nyi.mail.srv.osa (Postfix) with ESMTP id 359FE20BFF for ; Sat, 26 Nov 2011 10:19:24 -0500 (EST) Content-Disposition: inline In-Reply-To: <20111126111247.43cb5d35@lxorguk.ukuu.org.uk> Sender: linux-omap-owner@vger.kernel.org List-Id: linux-omap@vger.kernel.org To: Alan Cox , tom.leiming@gmail.com, tony@atomide.com, arnd@arndb.de, linux-omap@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, Ming Lei On Sat, Nov 26, 2011 at 11:12:47AM +0000, Alan Cox wrote: > On Sat, 26 Nov 2011 12:31:44 +0800 > tom.leiming@gmail.com wrote: > > > From: Ming Lei > > > > One face detection IP[1] is integared inside OMAP4 SoC, so > > introduce this driver to make face detection function work > > on OMAP4 SoC. > > > > This driver is platform independent, so in theory can > > be used to drive same IP module on other platforms. > > > > [1], ch9 of OMAP4 TRM > > > > Signed-off-by: Ming Lei > > If you are submitting it then it ought to have your sign off too. This > looks hardly ready for submission however. > > > +config FDIF > > + tristate "Face Detection module" > > + help > > + The FDIF is a face detection module, which can be integrated into > > + SoCs to detect the location of human beings' face in one image. At > > + least now, TI OMAP4 has the module inside. > > So we have a completelt device specific API for what is becoming a more > general feature and on some hardware is tied to camera and the like. This > seems wrong > > IMHO this should be part of video4linux because that'll make the > integrated stuff work sanely and for your cases it still provided the > right kind of mmap and format interfaces you need. I agree, as-is, this isn't ok, sorry, please work with the v4l developers on this. greg k-h