From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755430Ab1ACQYK (ORCPT ); Mon, 3 Jan 2011 11:24:10 -0500 Received: from moutng.kundenserver.de ([212.227.126.187]:61665 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754197Ab1ACQYI (ORCPT ); Mon, 3 Jan 2011 11:24:08 -0500 Date: Mon, 3 Jan 2011 17:24:02 +0100 (CET) From: Guennadi Liakhovetski X-X-Sender: lyakh@axis700.grange To: Alberto Panizzo cc: Mauro Carvalho Chehab , Hans Verkuil , Laurent Pinchart , Magnus Damm , =?ISO-8859-1?Q?M=E1rton_N=E9meth?= , linux-media@vger.kernel.org, linux-kernel Subject: Re: [PATCH 2/3] mx3_camera: Support correctly the YUV222 and BAYER configurations of CSI In-Reply-To: <1294070779.2493.53.camel@realization> Message-ID: References: <1290964687.3016.5.camel@realization> <1290965045.3016.11.camel@realization> <1294070779.2493.53.camel@realization> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Provags-ID: V02:K0:bOs8OYgDWgcRGEGNTg/QQb3OTfqMxMNuzMbIU2+29t6 vmy6ZFjsMmR8SWu5lhk7s8Rvt+ATzZWnKJXR+mPWxRZ5kHTKdy KZQSG86ZxbnqagZeoSfhTrxntXptOLnk/FUx3B149cqPZ+fwwU iiU+g/Gb6V47iEjDcCu7fVi8x8EBoRJQhRXqpERdvtCY95UbYx mpKYfPko169kXVWKCljZvAwF0CIqukK5WUi1pVqUMk= Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 3 Jan 2011, Alberto Panizzo wrote: [snip] > Guennadi, how do you consider the path I've shown? Can I continue in this > way or shall I present a patch that get rid of translations and manage > all the pixel format in the way I understood the manual speak about? > > I prefer this way that maintain the usability of the whole set of pixel codes > for everyone after this step and then fix together the translations. > I don't hold a camera that output a grey format.. Sorry, I'm not sure I understand what exactly you're proposing. Please, just look at my last reply, in which I explain, why I don't think this is a good way to fix things. You don't need to support various formats natively on CSI / IPU to be able to just pass data 1-to-1. If that is your only purpose, please, test it that way, and if it is broken, please, fix it. This would be good even if you actually want to support formats natively eventually, because that would also fix other formats with similar sample-per-pixel values. When this is fixed, if you want to actually support formats natively to perform some hardware-assisted format conversions, that also can be done, but then, please, explain this clearly in your patch (series). Thanks Guennadi --- Guennadi Liakhovetski, Ph.D. Freelance Open-Source Software Developer http://www.open-technology.de/