From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-qv1-f54.google.com (mail-qv1-f54.google.com [209.85.219.54]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id EC5742C9C for ; Wed, 5 Jan 2022 12:26:42 +0000 (UTC) Received: by mail-qv1-f54.google.com with SMTP id fo11so37338511qvb.4 for ; Wed, 05 Jan 2022 04:26:42 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=vanguardiasur-com-ar.20210112.gappssmtp.com; s=20210112; h=date:from:to:cc:subject:message-id:references:mime-version :content-disposition:in-reply-to; bh=+6zUFKikZ5uI7X+WBRedpvpkwDXnTgkeRYffGhJplBQ=; b=v0GeznWOJp1intYhv38FsyRKC2LY+s7ndDGtu8347+i9nbXXv04HckdcZQOYkPKp90 /jX/cn0VqKe2GUgFg8xH4/RRsX2gOU2vi4YIq2jnAXgtJFu8NKWOmv5UTm7JKOBG3wVd yX9NM8tByP/uwWU3/WPQ7QTsUzFb47l90lqhxnFgpjcN80pQu3wHho0EGORmw4jD4Hze 2N7+JCrk4rGAxSTsjSt08zwMWj9DQKe79mw7iOhhxX8p6AUqdjfORxMRC/f4QgTqsNJB wYnJmPOreCRXYJ/LqiEG5E0tR+gwvKZVim6Fcfba4VyFEW0Zya6GV1LFeZO/aFjf05Op DDbw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:date:from:to:cc:subject:message-id:references :mime-version:content-disposition:in-reply-to; bh=+6zUFKikZ5uI7X+WBRedpvpkwDXnTgkeRYffGhJplBQ=; b=olWRJbtf37p8P0blH0nhf76VIYM5ZarEDm1aVygDVmCQgtNQKVYosVugclIl1lHSiJ 6h69kWkAjB+/aDgcqOrZeZSzmP9cJEWquKUnelNVMglkeZtGRdf92VOeNQ2YGjF/W+IS F+2Q4ZQm++3yUBJJ/a2ssJnciWKVKPhPyKmCq5HQygowj3T6ksnZB6htpaJ6c6TLc40U GAH5g0A5hetbt70HkFQQR9L4Pi1Id9pc87QwEe/E+dIEkUY8rjOceLoqUnRmwjllM0Gj 7eOW7loNvpMGWkn96jCdsunTPReS8vsiwkA8FVAzWZHIORQ6sZMQG2oJqqV7YfeICc0g SSlA== X-Gm-Message-State: AOAM533hY/WHnYwgXC12Sme2IGk9WrQ4Ro4D2nlnuozqdcJ2Xf8ln3DY Hrh7M6yd6AMyuAlcWOlqaf/njw== X-Google-Smtp-Source: ABdhPJxqbVB7XWk6LPZe9NdWapRfu4IObMFem/esLdwpybt5KuvKvaxSKJK/v6LXDUp/1xX65da8KA== X-Received: by 2002:a05:6214:262d:: with SMTP id gv13mr49591568qvb.23.1641385601790; Wed, 05 Jan 2022 04:26:41 -0800 (PST) Received: from eze-laptop ([186.122.18.30]) by smtp.gmail.com with ESMTPSA id a15sm36481925qtb.5.2022.01.05.04.26.36 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 05 Jan 2022 04:26:39 -0800 (PST) Date: Wed, 5 Jan 2022 09:26:33 -0300 From: Ezequiel Garcia To: Benjamin Gaignard Cc: mchehab@kernel.org, p.zabel@pengutronix.de, gregkh@linuxfoundation.org, hverkuil-cisco@xs4all.nl, jernej.skrabec@gmail.com, nicolas.dufresne@collabora.co.uk, linux-media@vger.kernel.org, linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-staging@lists.linux.dev, kernel@collabora.com Subject: Re: [PATCH v4 1/2] media: hevc: Remove RPS named flags Message-ID: References: <20220104073842.1791639-1-benjamin.gaignard@collabora.com> <20220104073842.1791639-2-benjamin.gaignard@collabora.com> Precedence: bulk X-Mailing-List: linux-staging@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20220104073842.1791639-2-benjamin.gaignard@collabora.com> Hi Benjamin, On Tue, Jan 04, 2022 at 08:38:41AM +0100, Benjamin Gaignard wrote: > Marking a picture as long-term reference is valid for DPB but not for RPS. > Change flag name to match with it description in HEVC spec chapiter > "8.3.2 Decoding process for reference picture set". > Remove the other unused RPS flags. > A change like this, which is affecting a kernel interface and has impact on userspace and drivers, requires a lot more attention in the commit description. If I understand correctly, this change is related to how HEVC was first introduced and how the DPB was originally represented in V4L2. Originally, a DPB entry struct v4l2_hevc_dpb_entry had an rps field which can be: V4L2_HEVC_DPB_ENTRY_RPS_ST_CURR_BEFORE V4L2_HEVC_DPB_ENTRY_RPS_ST_CURR_AFTER V4L2_HEVC_DPB_ENTRY_RPS_LT_CURR Perhaps this idea followed libva, where a VAPictureHEVC has flags field which can be: VA_PICTURE_HEVC_RPS_ST_CURR_BEFORE, VA_PICTURE_HEVC_RPS_ST_CURR_AFTER, VA_PICTURE_HEVC_RPS_LT_CURR, VA_PICTURE_HEVC_LONG_TERM_REFERENCE In this representation, the sets PocStCurrBefore, PocStCurrAfter, PocLtCurr are implicit, and must be built by the kernel from the DPB entries struct v4l2_hevc_dpb_entry, using the information in the rps field. Last year, we changed this with your commit: commit d395a78db9eabd12633b39e05c80e803543b6590 Author: Benjamin Gaignard Date: Thu Jun 3 13:49:57 2021 +0200 media: hevc: Add decode params control Which added the control v4l2_ctrl_hevc_decode_params, which includes the sets PocStCurrBefore, PocStCurrAfter, PocLtCurr explicitly and therefore moved the responsability of creating and maintaining the sets to userspace. This effectively made the rps field in the struct v4l2_hevc_dpb_entry useless. The longterm flag is still needed though for a DPB entry. With this in mind, you could even say this commit is doing actually two things: 1. Removes the unused rps field. 2. Adds a flag field, for the longterm DPB entry boolean. Do you think a single byte of flags will be OK for a DPB entry? I think so, but I'd love yours/others input here. If the above is more or less accurate then, and provided you submit a new version with a more detailed commit description: Reviewed-by: Ezequiel Garcia Also, a small question: > Signed-off-by: Benjamin Gaignard > --- > version 4: > - check flags with & and not == > > Documentation/userspace-api/media/v4l/ext-ctrls-codec.rst | 8 +++----- > drivers/staging/media/hantro/hantro_g2_hevc_dec.c | 2 +- > drivers/staging/media/sunxi/cedrus/cedrus_h265.c | 2 +- > include/media/hevc-ctrls.h | 6 ++---- > 4 files changed, 7 insertions(+), 11 deletions(-) > > diff --git a/Documentation/userspace-api/media/v4l/ext-ctrls-codec.rst b/Documentation/userspace-api/media/v4l/ext-ctrls-codec.rst > index e141f0e4eec9..38da33e61c3d 100644 > --- a/Documentation/userspace-api/media/v4l/ext-ctrls-codec.rst > +++ b/Documentation/userspace-api/media/v4l/ext-ctrls-codec.rst > @@ -3166,11 +3166,9 @@ enum v4l2_mpeg_video_hevc_size_of_length_field - > :c:func:`v4l2_timeval_to_ns()` function to convert the struct > :c:type:`timeval` in struct :c:type:`v4l2_buffer` to a __u64. > * - __u8 > - - ``rps`` > - - The reference set for the reference frame > - (V4L2_HEVC_DPB_ENTRY_RPS_ST_CURR_BEFORE, > - V4L2_HEVC_DPB_ENTRY_RPS_ST_CURR_AFTER or > - V4L2_HEVC_DPB_ENTRY_RPS_LT_CURR) > + - ``flags`` > + - Long term flag for the reference frame > + (V4L2_HEVC_DPB_ENTRY_LONG_TERM_REFERENCE) Is this longterm flag associated in any way to a syntax element or an HEVC process? If so, please document that. Thanks, Ezequiel