From patchwork Thu May 13 15:21:08 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jacopo Mondi X-Patchwork-Id: 12271 Return-Path: X-Original-To: parsemail@patchwork.libcamera.org Delivered-To: parsemail@patchwork.libcamera.org Received: from lancelot.ideasonboard.com (lancelot.ideasonboard.com [92.243.16.209]) by patchwork.libcamera.org (Postfix) with ESMTPS id 794E7C31F1 for ; Thu, 13 May 2021 15:20:38 +0000 (UTC) Received: from lancelot.ideasonboard.com (localhost [IPv6:::1]) by lancelot.ideasonboard.com (Postfix) with ESMTP id BBF796891C; Thu, 13 May 2021 17:20:37 +0200 (CEST) Received: from relay6-d.mail.gandi.net (relay6-d.mail.gandi.net [217.70.183.198]) by lancelot.ideasonboard.com (Postfix) with ESMTPS id A568C688E4 for ; Thu, 13 May 2021 17:20:36 +0200 (CEST) X-Originating-IP: 79.53.131.195 Received: from uno.homenet.telecomitalia.it (host-79-53-131-195.retail.telecomitalia.it [79.53.131.195]) (Authenticated sender: jacopo@jmondi.org) by relay6-d.mail.gandi.net (Postfix) with ESMTPSA id C8669C0005; Thu, 13 May 2021 15:20:35 +0000 (UTC) From: Jacopo Mondi To: libcamera-devel@lists.libcamera.org Date: Thu, 13 May 2021 17:21:08 +0200 Message-Id: <20210513152116.17666-1-jacopo@jmondi.org> X-Mailer: git-send-email 2.31.1 MIME-Version: 1.0 Subject: [libcamera-devel] [PATCH v3 0/8] ipu3: imgu: Improve ImgU calculation procedure X-BeenThere: libcamera-devel@lists.libcamera.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: libcamera-devel-bounces@lists.libcamera.org Sender: "libcamera-devel" This series applies https://github.com/intel/intel-ipu3-pipecfg/commit/243d13446e44646c92118c0f0cb7fdc89b91f85c To the ImgU pipeline calculation. v3: Huston, we have problems, several ones. Two have them have a bug/issue associated, the third is run-time performace issue 1) https://github.com/intel/intel-ipu3-pipecfg/issues/2 2) https://bugs.libcamera.org/show_bug.cgi?id=32 3) patch 1/8 considerably increases the number of tested configurations Has you know the procedure is iterative. The script constructs a list of possible pipe configurations, then walks all of them to find the one with the larger FOV. Splendid! I took some measurement when starting the Android Camera HAL, which tests camera resolutions at startup time - currently with libcamera mailine we test 44718 combinations. The startup delay is barely noticeable - with this series applied, due to 3), 282340 combinations are tested. Starting the HAL is a bit slower, but nothing crazy. Also the HAL is usually started at system boot, so users are not impacted. - if I fix 2), as reported in the bug, 3644086 combinations are tested. Starting the HAL takes 5/10 seconds - if 1) gets fixed starting the HAL takes more than 15 seconds I think this is another hint that the current ImgU configuration procedure should be trashed as soon as possible, and the vendor should provide enough documentation/support to implement the pipe configuration in a sane way if this has to go in any production device for real. Jacopo Mondi (8): libcamera: ipu3: imgu: Update BDS calculation process libcamera: ipu3: imgu: Filter resolutions < IF_CROP_MAX libcamera: ipu3: imgu: Filter BDS by height libcamera: ipu3: imgu: Fix BDS height calculation libcamera: ipu3: imgu: Fix IF height selection libcamera: ipu3: imgu: Fix BSD height size comparison libcamera: ipu3: imgu: Add pipe calculation debug libcamera: ipu3: imgu: Bump script version src/libcamera/pipeline/ipu3/imgu.cpp | 117 ++++++++++++++++++++++----- 1 file changed, 95 insertions(+), 22 deletions(-) --- 2.31.1