Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 37633C6FD18 for ; Tue, 25 Apr 2023 11:30:39 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S233936AbjDYLah (ORCPT ); Tue, 25 Apr 2023 07:30:37 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40476 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S233903AbjDYLag (ORCPT ); Tue, 25 Apr 2023 07:30:36 -0400 Received: from mga18.intel.com (mga18.intel.com [134.134.136.126]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id EE4DC3C15 for ; Tue, 25 Apr 2023 04:30:34 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1682422235; x=1713958235; h=date:from:to:cc:subject:message-id:references: mime-version:in-reply-to; bh=214pIIsWl80owSPIKrQBtwRUJHUwPyVnBUN1uYek/Qc=; b=HfQkoTOpFXFoYQ5BI9aK8uAcOlVxbpbtgR4qVN26YxBJH3ArSgffsJqb UcD2iZLKpQ2fL8Vi79gzn+6y/qR0uI5M02WGau5WCBYSqZ144LzmXeLI3 q7Go84U+LwHVeDX+QZDp3rSNqPVwneJMw/j5FnK9BmdZl0x+EO2JdcMOl 24YoqFPz0LdzQqnysbSiTHE4nF29Q1LitSJ9q7sjroj414B/TrIMZuNEF qwwrxW5O4SdSMqZHLw+kLtQnWfXKhS7XZXItrdApSqtJtlynhdFfwDmvu 2+aemCnZWkaNsAcvne8pwZadTEXfjahua2QUIEGk9KquLBB+dCMqHTl7Q w==; X-IronPort-AV: E=McAfee;i="6600,9927,10690"; a="330950511" X-IronPort-AV: E=Sophos;i="5.99,225,1677571200"; d="scan'208";a="330950511" Received: from fmsmga007.fm.intel.com ([10.253.24.52]) by orsmga106.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 25 Apr 2023 04:30:33 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10690"; a="696154782" X-IronPort-AV: E=Sophos;i="5.99,225,1677571200"; d="scan'208";a="696154782" Received: from turnipsi.fi.intel.com (HELO kekkonen.fi.intel.com) ([10.237.72.44]) by fmsmga007-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 25 Apr 2023 04:30:30 -0700 Received: from kekkonen.localdomain (localhost [IPv6:::1]) by kekkonen.fi.intel.com (Postfix) with SMTP id D3E3811FAD0; Tue, 25 Apr 2023 14:30:27 +0300 (EEST) Date: Tue, 25 Apr 2023 14:30:27 +0300 From: Sakari Ailus To: Laurent Pinchart Cc: bingbu.cao@intel.com, linux-media@vger.kernel.org, ilpo.jarvinen@linux.intel.com, tfiga@chromium.org, senozhatsky@chromium.org, hdegoede@redhat.com, bingbu.cao@linux.intel.com, tian.shu.qiu@intel.com, hongju.wang@intel.com, daniel.h.kang@intel.com Subject: Re: [RFC PATCH 13/14] Documentation: add Intel IPU6 ISYS driver admin-guide doc Message-ID: References: <20230413100429.919622-1-bingbu.cao@intel.com> <20230413100429.919622-14-bingbu.cao@intel.com> <20230420144917.GD17497@pendragon.ideasonboard.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20230420144917.GD17497@pendragon.ideasonboard.com> Precedence: bulk List-ID: X-Mailing-List: linux-media@vger.kernel.org Hi Laurent, On Thu, Apr 20, 2023 at 05:49:17PM +0300, Laurent Pinchart wrote: ... > > +common driver which does PCI configuration, firmware loading and parsing, > > +firmware authentication, DMA mapping and IPU-MMU (internal Memory mapping Unit) > > +configuration. intel_ipu6_isys implements V4L2, Media Controller and V4L2 > > +sub-device interfaces. The IPU6 ISYS driver supports camera sensors connected > > +to the IPU6 ISYS through V4L2 sub-device sensor drivers. > > Will the intel_ipu6 driver also serve a future intel_ipu6_psys driver, > or is it specific to the ISYS ? intel-ipu6 contains common parts required by both ISYS and PSYS drivers. > > > + > > +Input system driver > > +=================== > > + > > +Input System driver mainly configure CSI2 DPHY, construct the firmware stream > > s/^Input/The Input/ > s/configure/configures the/ > s/construct/constructs/ > > > +configuration and send commands to firmware and get response from hardware and > > s/send/sends/ > s/get response/gets responses/ > > > +firmware and then return buffers to user. > > s/return/returns/ > > So control of the ISYS goes through the firmware, the driver doesn't > access the hardware directly ? Must of the low-level hardware control takes place through firmware, the ISYS driver uses direct register access mainly for the communication with the ISYS firmware and PHY configuration. > > > +The ISYS is represented as several V4L2 sub-devices - 'Intel IPU6 CSI2 $port', > > +which provide V4L2 subdev interfaces to the user space, there are also several > > +video nodes for each CSI-2 stream capture - 'Intel IPU6 ISYS capture $num' which > > +provide interface to user to set formats, queue buffers and streaming. > > + > > +.. kernel-figure:: ipu6_isys_graph.svg > > + :alt: ipu6 isys media graph > > I see in the figure 6 CSI-2 receivers and 4 video nodes. Are those CSI-2 > receivers independent of each other, or do they share SoCs pins ? For > instance, I'm thinking about a use case where you would have 4 data > lanes and 2 clock lanes, and support capturing from 2x 2-lanes sensors > with 2 CSI-2 RX, or 1x 4-lanes sensor with 1 CSI-2 RX. More generally, > can you document the input options of the ISYS ? For DT-based systems > I'd expect this information to be in the DT bindings, here it needs to > be documented separately. This is a to-do item (as noted on the cover page). The number of video nodes is also going to change with streams support. > > On the video device node side, are there 4 independent hardware DMA > engines that can be dynamically connected to the different CSI-2 > receivers, or is that what the current firmware version provide, with > the hardware able to capture more than 4 streams with firmware > modifications ? > > > + > > +Capturing frames by IPU6 ISYS > > +------------------------------------ > > + > > +IPU6 ISYS is used to capture frames from the camera sensors connected to the > > +CSI2 port. The supported input formats of ISYS are listed in table below: > > s/port/ports/ > > > + > > +.. tabularcolumns:: |p{0.8cm}|p{4.0cm}|p{4.0cm}| > > + > > +.. flat-table:: > > + :header-rows: 1 > > + > > + * - IPU6 ISYS supported input formats > > + > > + * - RGB565, RGB888 > > + > > + * - UYVY8, YUYV8 > > Unless I'm mistaken, CSI-2 doesn't have separate UYVY and YUYV formats. > Am I missing something ? > > > + > > + * - RAW8, RAW10, RAW12 > > Is embedded data capture supported ? More generically, can we capture > different data types and/or virtual channels from the same CSI-2 input ? Yes but not quite yet (streams support is pending). > > What's the maximum resolution supported by the ISYS ? Are the DMA > engines independent, or would the maximum resolution be different if you > capture from a single camera sensor or multiple camera sensors at the > same time ? My understanding is they're independent, i.e. the maximum resolution isn't affected by what else is happening in the device. -- Regards, Sakari Ailus