From patchwork Thu Jul 17 12:59:20 2025 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Stefan Klug X-Patchwork-Id: 23838 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 4D42CC3237 for ; Thu, 17 Jul 2025 12:59:39 +0000 (UTC) Received: from lancelot.ideasonboard.com (localhost [IPv6:::1]) by lancelot.ideasonboard.com (Postfix) with ESMTP id 2A80161517; Thu, 17 Jul 2025 14:59:38 +0200 (CEST) Authentication-Results: lancelot.ideasonboard.com; dkim=pass (1024-bit key; unprotected) header.d=ideasonboard.com header.i=@ideasonboard.com header.b="tSw3AhXT"; dkim-atps=neutral Received: from perceval.ideasonboard.com (perceval.ideasonboard.com [IPv6:2001:4b98:dc2:55:216:3eff:fef7:d647]) by lancelot.ideasonboard.com (Postfix) with ESMTPS id 6FC8E61517 for ; Thu, 17 Jul 2025 14:59:36 +0200 (CEST) Received: from ideasonboard.com (unknown [IPv6:2a00:6020:448c:6c00:7b93:8acd:d82d:248d]) by perceval.ideasonboard.com (Postfix) with UTF8SMTPSA id 2AC091FA4; Thu, 17 Jul 2025 14:59:02 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=ideasonboard.com; s=mail; t=1752757142; bh=vm9yni8W06P5krTo8HhITQr1z3umvQnBgtuVtwU9DaM=; h=From:To:Cc:Subject:Date:From; b=tSw3AhXTiZislvN3q0nFNU2EHwEorJ1vBFRjkgNpRNYUWsBoxPh4udN1RTTRpbYKp EGaMbePtwHT6EIwE0fEyUuFZrF8PfAThGrXO0YUlncsWfihkJVl3v6b7WpZ7dkDw6L fzTJrKwTcISuA9OBruYbA+mpTLLC7A+w2RPep4G4= From: Stefan Klug To: libcamera-devel@lists.libcamera.org Cc: Stefan Klug Subject: [PATCH v3 0/5] rkisp1: Allow usage of more than 4 buffers Date: Thu, 17 Jul 2025 14:59:20 +0200 Message-ID: <20250717125931.2848300-1-stefan.klug@ideasonboard.com> X-Mailer: git-send-email 2.48.1 MIME-Version: 1.0 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" Hi all, In the rkisp1 pipeline the value of StreamConfiuration.bufferCount is reset to 4 on validate(). This effectively limits the number of buffers that are in the loop to 4. As soon as the consumers (gstream/pipewire...) of the buffers need more than 1 buffer at a time, we start to see frame drops. Sven resurrected an old series [1] that solves the issue at the expense of an additional control and breaking changes in the API. In the IoB code camp two weeks ago there were some discussions regarding improvements in the rkisp1 regulation and per-frame-control (PFC). PFC has been an ongoing topic for quite some time now and it seems to slowly converge. Still we were hesitant in breaking the API or introducing new controls before the overall concepts have stabilized enough. To mitigate the pain mentioned above we came up with a minimal changeset that allows to request more than 4 buffers. Only 4 (or the pipeline depth reported by the pipeline) buffers are queued into the device. This has the same benefits as [1] but doesn't require an API break. Updates in Version 1 of the series (after the RFC): - Included the feedback from reviews of the RFC (For etails see the chaneglog in the individual patches) - Included two patches with fixes from Sven an Kieran - I modified patch (6/6) from Kieran a bit after the review comment from Laurent. @Kieran could you have a look at that as I don't have a easy way to replicate the issue? Ideally I would like to merge that patch into patch 2/6 as it fixes an issue introduced there. For now I left it separate for easier review. Updates in v2: - Fixed segfault due to wrong lambda capturing - Fixed ASSERT dues to recursive call of doQueueRequests in virtual pipeline - Improved documentation - Squashed patch 6 into patch 2 Updates in v3: - Split kPipelineDepth into kRkISP1MaxQueuedRequests and kRkISP1MinBufferCount. These contain the same value but convey the intent better and make the code clearer. - Ensure the bufferCount is reset to at least 4 in validate. - Fixed the underallocation of V4L2 buffer structs as proposed by Kieran - Incorporated some stylistic suggestions from review Best regards, Stefan Nícolas F. R. A. Prado (1): libcamera: pipeline: rkisp1: Don't rely on bufferCount Stefan Klug (4): libcamera: pipeline_handler: Move waitingRequests_ into camera class libcamera: pipeline_handler: Allow to limit the number of queued requests pipeline: rkisp1: Limit the maximum number of buffers queued in pipeline: rkisp1: Properly handle the bufferCount set in the stream configuration include/libcamera/internal/camera.h | 2 + include/libcamera/internal/pipeline_handler.h | 9 ++- src/libcamera/camera.cpp | 9 +++ src/libcamera/pipeline/rkisp1/rkisp1.cpp | 42 ++++++++--- src/libcamera/pipeline/rkisp1/rkisp1_path.cpp | 7 +- src/libcamera/pipeline/rkisp1/rkisp1_path.h | 4 +- src/libcamera/pipeline_handler.cpp | 72 +++++++++++++++---- 7 files changed, 106 insertions(+), 39 deletions(-)