[FFmpeg-devel] [PATCH] examples/demuxing_decoding: use correct size of video_dst_data
andreas.cadhalpun at googlemail.com
Tue May 12 14:32:54 CEST 2015
On 12.05.2015 11:17, wm4 wrote:
> On Mon, 11 May 2015 23:34:32 +0200
> Andreas Cadhalpun <andreas.cadhalpun at googlemail.com> wrote:
>> @@ -109,7 +108,7 @@ static int decode_packet(int *got_frame, int cached)
>> pix_fmt, width, height);
>> /* write to rawvideo file */
>> - fwrite(video_dst_data, 1, video_dst_bufsize, video_dst_file);
>> + fwrite(video_dst_data, 1, video_dst_linesize * height, video_dst_file);
> This would still write the line padding to the file, wouldn't it?
It wouldn't, because the video_dst_data buffer has no line padding.
(That's the only purpose of this buffer, see the comment above av_image_copy.)
> means it wouldn't write the full frame, _and_ each line would have
> garbage at the end of it.
> (Also I can't make sense of these examples...)
This one is quite useful for fuzzing, because it has much less overhead than ffmpeg.
More information about the ffmpeg-devel