From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-3.0 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,SPF_PASS,URIBL_BLOCKED,USER_AGENT_NEOMUTT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3F43BC4360F for ; Tue, 26 Mar 2019 13:54:48 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 1386A20823 for ; Tue, 26 Mar 2019 13:54:48 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1731681AbfCZNyr (ORCPT ); Tue, 26 Mar 2019 09:54:47 -0400 Received: from mga12.intel.com ([192.55.52.136]:49562 "EHLO mga12.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726175AbfCZNyq (ORCPT ); Tue, 26 Mar 2019 09:54:46 -0400 X-Amp-Result: UNKNOWN X-Amp-Original-Verdict: FILE UNKNOWN X-Amp-File-Uploaded: False Received: from orsmga004.jf.intel.com ([10.7.209.38]) by fmsmga106.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 26 Mar 2019 06:54:45 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.60,271,1549958400"; d="scan'208";a="286006541" Received: from paasikivi.fi.intel.com ([10.237.72.42]) by orsmga004.jf.intel.com with ESMTP; 26 Mar 2019 06:54:40 -0700 Received: by paasikivi.fi.intel.com (Postfix, from userid 1000) id C828A205E7; Tue, 26 Mar 2019 15:54:39 +0200 (EET) Date: Tue, 26 Mar 2019 15:54:39 +0200 From: Sakari Ailus To: Mickael GUENE Cc: "linux-media@vger.kernel.org" , Hugues FRUCHET , Mauro Carvalho Chehab , Matt Ranostay , Petr Cvek , Akinobu Mita , "linux-kernel@vger.kernel.org" , "David S. Miller" , Ben Kao , Nicolas Ferre , Todor Tomov , Rui Miguel Silva , Greg Kroah-Hartman , Hans Verkuil , Ricardo Ribalda Delgado , Jacopo Mondi , Tianshu Qiu , Bingbu Cao Subject: Re: [PATCH v3 2/2] media:st-mipid02: MIPID02 CSI-2 to PARALLEL bridge driver Message-ID: <20190326135439.oju2k6idndjaulfj@paasikivi.fi.intel.com> References: <1552373045-134493-1-git-send-email-mickael.guene@st.com> <1553594620-88280-1-git-send-email-mickael.guene@st.com> <1553594620-88280-3-git-send-email-mickael.guene@st.com> <20190326113308.mp5hqdw3ktbpcawu@paasikivi.fi.intel.com> <9c8fb935-5c67-d4aa-d0ba-2aae0edd2b55@st.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <9c8fb935-5c67-d4aa-d0ba-2aae0edd2b55@st.com> User-Agent: NeoMutt/20170113 (1.7.2) Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Mickael, On Tue, Mar 26, 2019 at 12:57:03PM +0000, Mickael GUENE wrote: ... > >> +static int mipid02_set_fmt(struct v4l2_subdev *sd, > >> + struct v4l2_subdev_pad_config *cfg, > >> + struct v4l2_subdev_format *format) > >> +{ > >> + struct v4l2_mbus_framefmt *mbus_fmt = &format->format; > >> + struct mipid02_dev *bridge = to_mipid02_dev(sd); > >> + struct i2c_client *client = bridge->i2c_client; > >> + struct v4l2_mbus_framefmt *fmt; > >> + int ret; > >> + > >> + dev_dbg(&client->dev, "%s for %d", __func__, format->pad); > >> + > >> + if (format->pad >= MIPID02_PAD_NB) > >> + return -EINVAL; > >> + /* second CSI-2 pad not yet supported */ > >> + if (format->pad == 1) > >> + return -EINVAL; > >> + > >> + mutex_lock(&bridge->lock); > >> + > >> + if (bridge->streaming) { > >> + ret = -EBUSY; > >> + goto error; > >> + } > >> + > >> + if (format->which == V4L2_SUBDEV_FORMAT_TRY) > >> + fmt = v4l2_subdev_get_try_format(sd, cfg, format->pad); > >> + else > >> + fmt = &bridge->fmt; > >> + > >> + *fmt = *mbus_fmt; > > > > What are the limits of the hardware regarding the size of the image? Aren't > > there any? > > > There are no limits for image size. > > The format on the sink pad needs to be propagated to the source pad as > > well. I presume there's nothing to set on the sink pad for this device, is > > there? > User only need to set format code so the driver can configure hardware. > In the mipid02 we have always the same format for sink and source. So I > only store one configuration during set_fmt when called from either pad0 > (sink) or pad2 (source). Is it the correct way to implement it ? or > should I only accept set_fmt on pad0 ? > For get_fmt I return stored configuration for pad0 and pad2. Only 76 or so characters per line, please. For pad 0 (sink) the format must be settable freely (as you don't have any hardware restrictions) and on the pad 2 (source) the driver converts the format set on the pad 0 according to the hardware functionality. This is what the link validation from the source pad onwards is based on. -- Regards, Sakari Ailus sakari.ailus@linux.intel.com