From: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
To: ffmpeg-devel@ffmpeg.org
Subject: Re: [FFmpeg-devel] [PATCH] avcodec/ffv1: Only allocate ThreadFrames for the decoder
Date: Tue, 6 Sep 2022 03:34:22 +0200
Message-ID: <GV1P250MB0737EA6E1A40D202D1E459588F7E9@GV1P250MB0737.EURP250.PROD.OUTLOOK.COM> (raw)
In-Reply-To: <AS8P250MB0744E83BB30F44AA2F7C05808F7C9@AS8P250MB0744.EURP250.PROD.OUTLOOK.COM>
Andreas Rheinhardt:
> The FFV1 decoder only uses the last frame's data to conceal
> errors. The encoder does not have this problem and therefore
> only uses the current frame and none of the ThreadFrames.
> So only allocate them for the decoder.
>
> Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
> ---
> libavcodec/ffv1.c | 13 -------------
> libavcodec/ffv1dec.c | 23 ++++++++++++++++++++++-
> 2 files changed, 22 insertions(+), 14 deletions(-)
>
> diff --git a/libavcodec/ffv1.c b/libavcodec/ffv1.c
> index c8781cdaaa..b6204740ed 100644
> --- a/libavcodec/ffv1.c
> +++ b/libavcodec/ffv1.c
> @@ -43,11 +43,6 @@ av_cold int ff_ffv1_common_init(AVCodecContext *avctx)
> s->avctx = avctx;
> s->flags = avctx->flags;
>
> - s->picture.f = av_frame_alloc();
> - s->last_picture.f = av_frame_alloc();
> - if (!s->picture.f || !s->last_picture.f)
> - return AVERROR(ENOMEM);
> -
> s->width = avctx->width;
> s->height = avctx->height;
>
> @@ -198,14 +193,6 @@ av_cold int ff_ffv1_close(AVCodecContext *avctx)
> FFV1Context *s = avctx->priv_data;
> int i, j;
>
> - if (s->picture.f)
> - ff_thread_release_ext_buffer(avctx, &s->picture);
> - av_frame_free(&s->picture.f);
> -
> - if (s->last_picture.f)
> - ff_thread_release_ext_buffer(avctx, &s->last_picture);
> - av_frame_free(&s->last_picture.f);
> -
> for (j = 0; j < s->max_slice_count; j++) {
> FFV1Context *fs = s->slice_context[j];
> for (i = 0; i < s->plane_count; i++) {
> diff --git a/libavcodec/ffv1dec.c b/libavcodec/ffv1dec.c
> index 794c58cc40..d4bc60a7da 100644
> --- a/libavcodec/ffv1dec.c
> +++ b/libavcodec/ffv1dec.c
> @@ -823,6 +823,11 @@ static av_cold int decode_init(AVCodecContext *avctx)
> if ((ret = ff_ffv1_common_init(avctx)) < 0)
> return ret;
>
> + f->picture.f = av_frame_alloc();
> + f->last_picture.f = av_frame_alloc();
> + if (!f->picture.f || !f->last_picture.f)
> + return AVERROR(ENOMEM);
> +
> if (avctx->extradata_size > 0 && (ret = read_extra_header(f)) < 0)
> return ret;
>
> @@ -1068,6 +1073,22 @@ static int update_thread_context(AVCodecContext *dst, const AVCodecContext *src)
> }
> #endif
>
> +static av_cold int ffv1_decode_close(AVCodecContext *avctx)
> +{
> + FFV1Context *const s = avctx->priv_data;
> +
> + if (s->picture.f) {
> + ff_thread_release_ext_buffer(avctx, &s->picture);
> + av_frame_free(&s->picture.f);
> + }
> +
> + if (s->last_picture.f) {
> + ff_thread_release_ext_buffer(avctx, &s->last_picture);
> + av_frame_free(&s->last_picture.f);
> + }
> + return ff_ffv1_close(avctx);
> +}
> +
> const FFCodec ff_ffv1_decoder = {
> .p.name = "ffv1",
> CODEC_LONG_NAME("FFmpeg video codec #1"),
> @@ -1075,7 +1096,7 @@ const FFCodec ff_ffv1_decoder = {
> .p.id = AV_CODEC_ID_FFV1,
> .priv_data_size = sizeof(FFV1Context),
> .init = decode_init,
> - .close = ff_ffv1_close,
> + .close = ffv1_decode_close,
> FF_CODEC_DECODE_CB(decode_frame),
> UPDATE_THREAD_CONTEXT(update_thread_context),
> .p.capabilities = AV_CODEC_CAP_DR1 /*| AV_CODEC_CAP_DRAW_HORIZ_BAND*/ |
Will apply this tomorrow unless there are objections.
- Andreas
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
prev parent reply other threads:[~2022-09-06 1:34 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-04 23:28 Andreas Rheinhardt
2022-09-06 1:34 ` Andreas Rheinhardt [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=GV1P250MB0737EA6E1A40D202D1E459588F7E9@GV1P250MB0737.EURP250.PROD.OUTLOOK.COM \
--to=andreas.rheinhardt@outlook.com \
--cc=ffmpeg-devel@ffmpeg.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
This inbox may be cloned and mirrored by anyone:
git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git
# If you have public-inbox 1.1+ installed, you may
# initialize and index your mirror using the following commands:
public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
ffmpegdev@gitmailbox.com
public-inbox-index ffmpegdev
Example config snippet for mirrors.
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git