From section 7.4.3.3 ("Decoded reference picture marking semantics") of the H.264 specifications: "The content of the decoded reference picture marking syntax structure shall be the same in all slice headers of the primary coded picture. When one or more redundant coded pictures are present, the content of the decoded reference picture marking syntax structure shall be the same in all slice headers of a redundant coded picture with a particular value of redundant_pic_cnt." And regarding redundant coded pictures: "[T]he content of the decoded reference picture marking syntax structure in a redundant coded picture is constrained in the way that the marking status of reference pictures and the value of frame_num after the decoded reference picture marking process in clause 8.2.5 must be identical regardless whether the primary coded picture or any redundant coded picture of the access unit would be decoded." So just looking at the first slice header is fine.
Signed-off-by: Andreas Rheinhardt <andreas.rheinha...@gmail.com> --- libavcodec/h264_parser.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/libavcodec/h264_parser.c b/libavcodec/h264_parser.c index 669a9a5ff7..604cddada5 100644 --- a/libavcodec/h264_parser.c +++ b/libavcodec/h264_parser.c @@ -468,10 +468,7 @@ static inline int parse_nal_units(AVCodecParserContext *s, if (ret < 0) goto fail; - /* Continue parsing to check if MMCO_RESET is present. - * FIXME: MMCO_RESET could appear in non-first slice. - * Maybe, we should parse all undisposable non-IDR slice of this - * picture until encountering MMCO_RESET in a slice of it. */ + /* Continue parsing to check if MMCO_RESET is present. */ if (nal.ref_idc && nal.type != H264_NAL_IDR_SLICE) { got_reset = scan_mmco_reset(s, &nal.gb, avctx); if (got_reset < 0) -- 2.21.0 _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".