[FFmpeg-devel] [PATCH v2 0/6] Implement SEI parsing for QSV decoders

1 Jun 2022

      Missing SEI information has always been a major drawback when using the QSV
decoders. I used to think that there's no chance to get at the data without
explicit implementation from the MSDK side (or doing something weird like
parsing in parallel). It turned out that there's a hardly known api method
that provides access to all SEI (h264/hevc) or user data (mpeg2video).

This allows to get things like closed captions, frame packing, display
orientation, HDR data (mastering display, content light level, etc.) without
having to rely on those data being provided by the MSDK as extended buffers.

The commit "Implement SEI parsing for QSV decoders" includes some hard-coded
workarounds for MSDK bugs which I reported:
https://github.com/Intel-Media-SDK/MediaSDK/issues/2597#issuecomment-1072795...

But that doesn't help. Those bugs exist and I'm sharing my workarounds,
which are empirically determined by testing a range of files. If someone is
interested, I can provide private access to a repository where we have been
testing this. Alternatively, I could also leave those workarounds out, and
just skip those SEI types.

In a previous version of this patchset, there was a concern that payload
data might need to be re-ordered. Meanwhile I have researched this carefully
and the conclusion is that this is not required.

My detailed analysis can be found here:
https://gist.github.com/softworkz/36c49586a8610813a32270ee3947a932

v2

 * qsvdec: make error handling consistent and clear
 * qsvdec: remove AV_CODEC_ID_MPEG1VIDEO constants
 * hevcdec: rename function to ff_hevc_set_side_data(), add doc text

softworkz (6):
  avutil/frame: Add av_frame_copy_side_data() and
    av_frame_remove_all_side_data()
  avcodec/vpp_qsv: Copy side data from input to output frame
  avcodec/mpeg12dec: make mpeg_decode_user_data() accessible
  avcodec/hevcdec: make set_side_data() accessible
  avcodec/h264dec: make h264_export_frame_props() accessible
  avcodec/qsvdec: Implement SEI parsing for QSV decoders

 doc/APIchanges               |   4 +
 libavcodec/h264_slice.c      |  98 ++++++++-------
 libavcodec/h264dec.h         |   2 +
 libavcodec/hevcdec.c         | 117 +++++++++---------
 libavcodec/hevcdec.h         |   9 ++
 libavcodec/mpeg12.h          |  28 +++++
 libavcodec/mpeg12dec.c       |  40 +-----
 libavcodec/qsvdec.c          | 234 +++++++++++++++++++++++++++++++++++
 libavfilter/qsvvpp.c         |   6 +
 libavfilter/vf_overlay_qsv.c |  19 ++-
 libavutil/frame.c            |  67 ++++++----
 libavutil/frame.h            |  32 +++++
 libavutil/version.h          |   2 +-
 13 files changed, 485 insertions(+), 173 deletions(-)

base-commit: b033913d1c5998a29dfd13e9906dd707ff6eff12
Published-As: https://github.com/ffstaging/FFmpeg/releases/tag/pr-ffstaging-31%2Fsoftworkz...
Fetch-It-Via: git fetch https://github.com/ffstaging/FFmpeg pr-ffstaging-31/softworkz/submit_qsv_sei-v2
Pull-Request: https://github.com/ffstaging/FFmpeg/pull/31

Range-diff vs v1:

 1:  4ee6cb47db = 1:  4ee6cb47db avutil/frame: Add av_frame_copy_side_data() and av_frame_remove_all_side_data()
 2:  3152156c97 = 2:  3152156c97 avcodec/vpp_qsv: Copy side data from input to output frame
 3:  8082c3ab84 = 3:  8082c3ab84 avcodec/mpeg12dec: make mpeg_decode_user_data() accessible
 4:  485d7f913d ! 4:  306bdaa39c avcodec/hevcdec: make set_side_data() accessible
     @@ libavcodec/hevcdec.c: error:
       }

      -static int set_side_data(HEVCContext *s)
     -+int ff_set_side_data(AVCodecContext *logctx, HEVCSEI *sei, HEVCContext *s, AVFrame *out)
     ++int ff_hevc_set_side_data(AVCodecContext *logctx, HEVCSEI *sei, HEVCContext *s, AVFrame *out)
       {
      -    AVFrame *out = s->ref->frame;
      -    int ret;
     @@ libavcodec/hevcdec.c: static int hevc_frame_start(HEVCContext *s)
           }

      -    ret = set_side_data(s);
     -+    ret = ff_set_side_data(s->avctx, &s->sei, s, s->ref->frame);
     ++    ret = ff_hevc_set_side_data(s->avctx, &s->sei, s, s->ref->frame);
           if (ret < 0)
               goto fail;

     @@ libavcodec/hevcdec.h: void ff_hevc_hls_residual_coding(HEVCContext *s, int x0, i

       void ff_hevc_hls_mvd_coding(HEVCContext *s, int x0, int y0, int log2_cb_size);

     -+int ff_set_side_data(AVCodecContext *logctx, HEVCSEI *sei, HEVCContext *s, AVFrame *out);
     ++/**
     ++ * Set the decodec side data to an AVFrame.
     ++ * @logctx context for logging.
     ++ * @sei HEVCSEI decoding context, must not be NULL.
     ++ * @s HEVCContext, can be NULL.
     ++ * @return < 0 on error, 0 otherwise.
     ++ */
     ++int ff_hevc_set_side_data(AVCodecContext *logctx, HEVCSEI *sei, HEVCContext *s, AVFrame *out);
      +
       extern const uint8_t ff_hevc_qpel_extra_before[4];
       extern const uint8_t ff_hevc_qpel_extra_after[4];
 5:  fb5c3df8e5 = 5:  16f5dfbfd1 avcodec/h264dec: make h264_export_frame_props() accessible
 6:  dcf08cd7b7 ! 6:  23de6d2774 avcodec/qsvdec: Implement SEI parsing for QSV decoders
     @@ libavcodec/qsvdec.c: static int qsv_export_film_grain(AVCodecContext *avctx, mfx
      +        memset(payload.Data, 0, payload.BufSize);
      +
      +        ret = MFXVideoDECODE_GetPayload(q->session, &ts, &payload);
     -+        if (ret != MFX_ERR_NONE) {
     -+            av_log(avctx, AV_LOG_WARNING, "error getting SEI payload: %d \n", ret);
     -+            return ret;
     ++        if (ret == MFX_ERR_NOT_ENOUGH_BUFFER) {
     ++            av_log(avctx, AV_LOG_WARNING, "Warning: Insufficient buffer on GetPayload(). Size: %"PRIu64" Needed: %d\n", sizeof(q->payload_buffer), payload.BufSize);
     ++            return 0;
      +        }
     ++        if (ret != MFX_ERR_NONE)
     ++            return ret;
      +
     -+        if (payload.NumBit == 0 || payload.NumBit >= payload.BufSize * 8) {
     ++        if (payload.NumBit == 0 || payload.NumBit >= payload.BufSize * 8)
      +            break;
     -+        }
      +
      +        start = find_start_offset(payload.Data);
      +
     @@ libavcodec/qsvdec.c: static int qsv_export_film_grain(AVCodecContext *avctx, mfx
      +        }
      +
      +        if (init_get_bits(&gb, &payload.Data[start], payload.NumBit - start * 8) < 0)
     -+            av_log(avctx, AV_LOG_ERROR, "Error initializing bitstream reader");
     -+        else
     ++            av_log(avctx, AV_LOG_ERROR, "Error initializing bitstream reader SEI type: %d  Numbits %d error: %d\n", payload.Type, payload.NumBit, ret);
     ++        else {
      +            ret = ff_h264_sei_decode(&sei, &gb, NULL, avctx);
      +
     -+        if (ret < 0)
     -+            av_log(avctx, AV_LOG_WARNING, "Error parsing SEI type: %d  Numbits %d error: %d\n", payload.Type, payload.NumBit, ret);
     -+        else
     -+            av_log(avctx, AV_LOG_DEBUG, "mfxPayload Type: %d  Numbits %d\n", payload.Type, payload.NumBit);
     ++            if (ret < 0)
     ++                av_log(avctx, AV_LOG_WARNING, "Failed to parse SEI type: %d  Numbits %d error: %d\n", payload.Type, payload.NumBit, ret);
     ++            else
     ++                av_log(avctx, AV_LOG_DEBUG, "mfxPayload Type: %d  Numbits %d\n", payload.Type, payload.NumBit);
     ++        }
      +    }
      +
      +    if (out)
     @@ libavcodec/qsvdec.c: static int qsv_export_film_grain(AVCodecContext *avctx, mfx
      +        memset(payload.Data, 0, payload.BufSize);
      +
      +        ret = MFXVideoDECODE_GetPayload(q->session, &ts, &payload);
     -+        if (ret != MFX_ERR_NONE) {
     -+            av_log(avctx, AV_LOG_WARNING, "error getting SEI payload: %d \n", ret);
     ++        if (ret == MFX_ERR_NOT_ENOUGH_BUFFER) {
     ++            av_log(avctx, AV_LOG_WARNING, "Warning: Insufficient buffer on GetPayload(). Size: %"PRIu64" Needed: %d\n", sizeof(q->payload_buffer), payload.BufSize);
      +            return 0;
      +        }
     ++        if (ret != MFX_ERR_NONE)
     ++            return ret;
      +
     -+        if (payload.NumBit == 0 || payload.NumBit >= payload.BufSize * 8) {
     ++        if (payload.NumBit == 0 || payload.NumBit >= payload.BufSize * 8)
      +            break;
     -+        }
      +
      +        if (!has_logged) {
      +            has_logged = 1;
     @@ libavcodec/qsvdec.c: static int qsv_export_film_grain(AVCodecContext *avctx, mfx
      +        }
      +
      +        if (init_get_bits(&gb, &payload.Data[start], payload.NumBit - start * 8) < 0)
     -+            av_log(avctx, AV_LOG_ERROR, "Error initializing bitstream reader");
     -+        else
     -+            ret = ff_hevc_decode_nal_sei(&gb, avctx, &sei, &ps, HEVC_NAL_SEI_PREFIX);
     ++            av_log(avctx, AV_LOG_ERROR, "Error initializing bitstream reader SEI type: %d  Numbits %d error: %d\n", payload.Type, payload.NumBit, ret);
     ++        else {
     ++            ret = ff_h264_sei_decode(&sei, &gb, NULL, avctx);
      +
     -+        if (ret < 0)
     -+            av_log(avctx, AV_LOG_WARNING, "error parsing SEI type: %d  Numbits %d error: %d\n", payload.Type, payload.NumBit, ret);
     -+        else
     -+            av_log(avctx, AV_LOG_DEBUG, "mfxPayload Type: %d  Numbits %d\n", payload.Type, payload.NumBit);
     ++            if (ret < 0)
     ++                av_log(avctx, AV_LOG_WARNING, "Failed to parse SEI type: %d  Numbits %d error: %d\n", payload.Type, payload.NumBit, ret);
     ++            else
     ++                av_log(avctx, AV_LOG_DEBUG, "mfxPayload Type: %d  Numbits %d\n", payload.Type, payload.NumBit);
     ++        }
      +    }
      +
      +    if (has_logged) {
     @@ libavcodec/qsvdec.c: static int qsv_export_film_grain(AVCodecContext *avctx, mfx
      +    }
      +
      +    if (out && out->frame)
     -+        return ff_set_side_data(avctx, &sei, NULL, out->frame);
     ++        return ff_hevc_set_side_data(avctx, &sei, NULL, out->frame);
      +
      +    return 0;
      +}
     @@ libavcodec/qsvdec.c: static int qsv_export_film_grain(AVCodecContext *avctx, mfx
      +
      +        memset(payload.Data, 0, payload.BufSize);
      +        ret = MFXVideoDECODE_GetPayload(q->session, &ts, &payload);
     -+        if (ret != MFX_ERR_NONE) {
     -+            av_log(avctx, AV_LOG_WARNING, "error getting SEI payload: %d \n", ret);
     -+            return ret;
     ++        if (ret == MFX_ERR_NOT_ENOUGH_BUFFER) {
     ++            av_log(avctx, AV_LOG_WARNING, "Warning: Insufficient buffer on GetPayload(). Size: %"PRIu64" Needed: %d\n", sizeof(q->payload_buffer), payload.BufSize);
     ++            return 0;
      +        }
     ++        if (ret != MFX_ERR_NONE)
     ++            return ret;
      +
     -+        if (payload.NumBit == 0 || payload.NumBit >= payload.BufSize * 8) {
     ++        if (payload.NumBit == 0 || payload.NumBit >= payload.BufSize * 8)
      +            break;
     -+        }
      +
      +        start = find_start_offset(payload.Data);
      +
     @@ libavcodec/qsvdec.c: static int qsv_export_film_grain(AVCodecContext *avctx, mfx
      +
      +        ff_mpeg_decode_user_data(avctx, mpeg_ctx, &payload.Data[start], (int)((payload.NumBit + 7) / 8) - start);
      +
     -+        if (ret < 0)
     -+            av_log(avctx, AV_LOG_WARNING, "error parsing SEI type: %d  Numbits %d error: %d\n", payload.Type, payload.NumBit, ret);
     -+        else
     -+            av_log(avctx, AV_LOG_DEBUG, "mfxPayload Type: %d  Numbits %d start %d -> %.s\n", payload.Type, payload.NumBit, start, (char *)(&payload.Data[start]));
     ++        av_log(avctx, AV_LOG_DEBUG, "mfxPayload Type: %d  Numbits %d start %d -> %.s\n", payload.Type, payload.NumBit, start, (char *)(&payload.Data[start]));
      +    }
      +
      +    if (!out)
     @@ libavcodec/qsvdec.c: static int qsv_decode(AVCodecContext *avctx, QSVContext *q,
                                                     insurf, &outsurf, sync);
               if (ret == MFX_WRN_DEVICE_BUSY)
                   av_usleep(500);
     -+        else if (avctx->codec_id == AV_CODEC_ID_MPEG1VIDEO || avctx->codec_id == AV_CODEC_ID_MPEG2VIDEO)
     ++        else if (avctx->codec_id == AV_CODEC_ID_MPEG2VIDEO)
      +            parse_sei_mpeg12(avctx, q, NULL);

           } while (ret == MFX_WRN_DEVICE_BUSY || ret == MFX_ERR_MORE_SURFACE);
     @@ libavcodec/qsvdec.c: static int qsv_decode(AVCodecContext *avctx, QSVContext *q,
               }

      +        switch (avctx->codec_id) {
     -+        case AV_CODEC_ID_MPEG1VIDEO:
      +        case AV_CODEC_ID_MPEG2VIDEO:
      +            ret = parse_sei_mpeg12(avctx, q, out_frame->frame);
      +            break;
     @@ libavcodec/qsvdec.c: static int qsv_decode(AVCodecContext *avctx, QSVContext *q,
      +        }
      +
      +        if (ret < 0)
     -+            av_log(avctx, AV_LOG_ERROR, "Error parsing SEI data\n");
     ++            av_log(avctx, AV_LOG_ERROR, "Error parsing SEI data: %d\n", ret);
      +
               out_frame->queued += 1;

-- 
ffmpeg-codebot

[FFmpeg-devel] [PATCH v2 0/6] Implement SEI parsing for QSV decoders

ffmpegagent