From: Anton Khirnov <anton@khirnov.net>
To: FFmpeg development discussions and patches <ffmpeg-devel@ffmpeg.org>
Subject: Re: [FFmpeg-devel] [PATCH v2] avutil/frame: add av_frame_replace
Date: Tue, 07 Jun 2022 12:21:32 +0200
Message-ID: <165459729289.13099.12969587257195843072@lain> (raw)
In-Reply-To: <20220605174458.1942-1-jamrial@gmail.com>
Quoting James Almer (2022-06-05 19:44:58)
> +
> + /* duplicate the frame data if it's not refcounted */
> + if (!src->buf[0]) {
> + for (int i = 0; i < FF_ARRAY_ELEMS(dst->buf); i++)
> + av_buffer_unref(&dst->buf[i]);
> + for (int i = 0; i < dst->nb_extended_buf; i++)
> + av_buffer_unref(&dst->extended_buf[i]);
> + av_freep(&dst->extended_buf);
> +
> + memset(dst->data, 0, sizeof(dst->data));
> + if (dst->extended_data != dst->data)
> + av_freep(&dst->extended_data);
> +
> + ret = av_frame_get_buffer(dst, 0);
> + if (ret < 0)
> + goto fail;
> +
> + ret = av_frame_copy(dst, src);
> + if (ret < 0)
> + goto fail;
> +
> + ret = av_buffer_replace(&dst->hw_frames_ctx, src->hw_frames_ctx);
> + if (ret < 0)
> + goto fail;
This looks suspicious, since we just allocated normal buffers for dst.
Also, can't t this whole block be replaced by just
if (!src->buf[0]) {
av_frame_unref(dst);
return av_frame_ref(dst, src);
}
at the beginning of this function?
> + if (src->extended_data != src->data) {
> + int ch = src->ch_layout.nb_channels;
> +
> +#if FF_API_OLD_CHANNEL_LAYOUT
> +FF_DISABLE_DEPRECATION_WARNINGS
> + if (!ch) {
> + ch = src->channels;
> + CHECK_CHANNELS_CONSISTENCY(src);
> + }
> +FF_ENABLE_DEPRECATION_WARNINGS
> +#endif
> + if (!ch) {
> + ret = AVERROR(EINVAL);
> + goto fail;
> + }
> +
> + dst->extended_data = av_malloc_array(sizeof(*dst->extended_data), ch);
> + if (!dst->extended_data) {
> + ret = AVERROR(ENOMEM);
> + goto fail;
> + }
> + memcpy(dst->extended_data, src->extended_data, sizeof(*src->extended_data) * ch);
nit: av_memdup()?
> + } else
> + dst->extended_data = dst->data;
> +
> + memcpy(dst->data, src->data, sizeof(src->data));
> + memcpy(dst->linesize, src->linesize, sizeof(src->linesize));
> +
> + return 0;
> +
> +fail:
> + av_frame_unref(dst);
> + return ret;
> +}
> +
> AVFrame *av_frame_clone(const AVFrame *src)
> {
> AVFrame *ret = av_frame_alloc();
> diff --git a/libavutil/frame.h b/libavutil/frame.h
> index 33fac2054c..e5c10e2b66 100644
> --- a/libavutil/frame.h
> +++ b/libavutil/frame.h
> @@ -752,6 +752,19 @@ void av_frame_free(AVFrame **frame);
> */
> int av_frame_ref(AVFrame *dst, const AVFrame *src);
>
> +/**
> + * Ensure the destination frame refers to the same data described by the source
> + * frame by creating a new reference for each AVBufferRef from src if they
> + * differ from those in dst, or if src is not reference counted, by allocating
> + * new buffers and copying data.
> + *
> + * Frame properties on dst will be replaced by those from src.
> + *
> + * @return 0 on success, a negative AVERROR on error. On error, dst is
> + * unreferenced.
> + */
> +int av_frame_replace(AVFrame *dst, const AVFrame *src);
An important property of av_buffer_replace() is that it's equivalent to
av_buffer_unref(dst) when src is NULL. It would probably be desirable
for av_frame_replace() to work the same way - currently it will try to
call av_frame_get_buffer() with invalid parameters and fail.
--
Anton Khirnov
_______________________________________________
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".
next prev parent reply other threads:[~2022-06-07 10:21 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-05 15:28 [FFmpeg-devel] [PATCH] " James Almer
2022-06-05 15:42 ` Andreas Rheinhardt
2022-06-05 15:52 ` James Almer
2022-06-05 17:44 ` [FFmpeg-devel] [PATCH v2] " James Almer
2022-06-07 10:21 ` Anton Khirnov [this message]
2022-06-07 11:39 ` James Almer
2022-06-07 11:47 ` Anton Khirnov
2022-06-07 12:02 ` [FFmpeg-devel] [PATCH v3] " James Almer
2022-06-07 12:18 ` Nicolas George
2022-06-07 12:22 ` James Almer
2022-06-07 12:23 ` Nicolas George
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=165459729289.13099.12969587257195843072@lain \
--to=anton@khirnov.net \
--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