November 2019 - ffmpeg-cvslog

lavc/rav1e: log and doc updated for const quantizer mode
by Zhong Li 28 Nov '19

28 Nov '19

ffmpeg | branch: master | Zhong Li <zhongli_dev(a)126.com> | Tue Nov 12 09:09:20 2019 +0000| [846e26b8c93a313d9f0c1f7d4a2965b357482abf] | committer: Zhong Li lavc/rav1e: log and doc updated for const quantizer mode Signed-off-by: Zhong Li <zhongli_dev(a)126.com> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=846e26b8c93a313d9… --- doc/encoders.texi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/doc/encoders.texi b/doc/encoders.texi index 6cf3a74093..6b02c7798d 100644 --- a/doc/encoders.texi +++ b/doc/encoders.texi @@ -1395,7 +1395,7 @@ Sets the maximum quantizer to use when using bitrate mode. Sets the minimum quantizer to use when using bitrate mode. @item qp -Uses quantizer mode to encode at the given quantizer. +Uses quantizer mode to encode at the given quantizer (0-255). @item speed Selects the speed preset (0-10) to encode with.

1 0

Add options for spatial layers.
by Thierry Foucu 28 Nov '19

28 Nov '19

ffmpeg | branch: master | Thierry Foucu <tfoucu(a)gmail.com> | Wed Nov 27 16:25:33 2019 -0800| [eaf566ce0b9fd33120cc564bc7cd6986de3ca724] | committer: James Almer Add options for spatial layers. Disable by default to output all the layers, to match libaomdec wrapper. Add option to select the operating point for the spatial layers. Update the documentation with the new options. Signed-off-by: James Almer <jamrial(a)gmail.com> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=eaf566ce0b9fd3312… --- doc/decoders.texi | 7 +++++++ libavcodec/libdav1d.c | 8 ++++++++ libavcodec/version.h | 2 +- 3 files changed, 16 insertions(+), 1 deletion(-) diff --git a/doc/decoders.texi b/doc/decoders.texi index 676e062e72..f18226b350 100644 --- a/doc/decoders.texi +++ b/doc/decoders.texi @@ -71,6 +71,13 @@ Set amount of tile threads to use during decoding. The default value is 0 (autod Apply film grain to the decoded video if present in the bitstream. Defaults to the internal default of the library. +@item oppoint +Select an operating point of a scalable AV1 bitstream (0 - 31). Defaults to the +internal default of the library. + +@item alllayers +Output all spatial layers of a scalable AV1 bitstream. The default value is false. + @end table @section libdavs2 diff --git a/libavcodec/libdav1d.c b/libavcodec/libdav1d.c index cf4b178f1d..5248e3f9f5 100644 --- a/libavcodec/libdav1d.c +++ b/libavcodec/libdav1d.c @@ -40,6 +40,8 @@ typedef struct Libdav1dContext { int tile_threads; int frame_threads; int apply_grain; + int operating_point; + int all_layers; } Libdav1dContext; static const enum AVPixelFormat pix_fmt[][3] = { @@ -134,6 +136,10 @@ static av_cold int libdav1d_init(AVCodecContext *c) if (dav1d->apply_grain >= 0) s.apply_grain = dav1d->apply_grain; + s.all_layers = dav1d->all_layers; + if (dav1d->operating_point >= 0) + s.operating_point = dav1d->operating_point; + s.n_tile_threads = dav1d->tile_threads ? dav1d->tile_threads : FFMIN(floor(sqrt(threads)), DAV1D_MAX_TILE_THREADS); @@ -378,6 +384,8 @@ static const AVOption libdav1d_options[] = { { "tilethreads", "Tile threads", OFFSET(tile_threads), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, DAV1D_MAX_TILE_THREADS, VD }, { "framethreads", "Frame threads", OFFSET(frame_threads), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, DAV1D_MAX_FRAME_THREADS, VD }, { "filmgrain", "Apply Film Grain", OFFSET(apply_grain), AV_OPT_TYPE_BOOL, { .i64 = -1 }, -1, 1, VD }, + { "oppoint", "Select an operating point of the scalable bitstream", OFFSET(operating_point), AV_OPT_TYPE_INT, { .i64 = -1 }, -1, 31, VD }, + { "alllayers", "Output all spatial layers", OFFSET(all_layers), AV_OPT_TYPE_BOOL, { .i64 = 0 }, 0, 1, VD }, { NULL } }; diff --git a/libavcodec/version.h b/libavcodec/version.h index efa9d743f7..8b9c27378c 100644 --- a/libavcodec/version.h +++ b/libavcodec/version.h @@ -29,7 +29,7 @@ #define LIBAVCODEC_VERSION_MAJOR 58 #define LIBAVCODEC_VERSION_MINOR 64 -#define LIBAVCODEC_VERSION_MICRO 100 +#define LIBAVCODEC_VERSION_MICRO 101 #define LIBAVCODEC_VERSION_INT AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \ LIBAVCODEC_VERSION_MINOR, \

1 0

avcodec: add mvha video decoder
by Paul B Mahol 27 Nov '19

27 Nov '19

ffmpeg | branch: master | Paul B Mahol <onemda(a)gmail.com> | Mon Nov 25 12:59:56 2019 +0100| [ed58f8475fa8e659abc887fcd4ef1bd886be397d] | committer: Paul B Mahol avcodec: add mvha video decoder > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=ed58f8475fa8e659a… --- Changelog | 1 + configure | 1 + libavcodec/Makefile | 1 + libavcodec/allcodecs.c | 1 + libavcodec/avcodec.h | 1 + libavcodec/codec_desc.c | 7 ++ libavcodec/mvha.c | 310 ++++++++++++++++++++++++++++++++++++++++++++++++ libavcodec/version.h | 2 +- libavformat/riff.c | 1 + 9 files changed, 324 insertions(+), 1 deletion(-) diff --git a/Changelog b/Changelog index 91d1b5a51e..f30f398a9f 100644 --- a/Changelog +++ b/Changelog @@ -26,6 +26,7 @@ version <next>: - AV1 Annex B demuxer - axcorrelate filter - mvdv decoder +- mvha decoder version 4.2: diff --git a/configure b/configure index 2519e6421f..ca7137f341 100755 --- a/configure +++ b/configure @@ -2779,6 +2779,7 @@ msmpeg4v3_decoder_select="h263_decoder" msmpeg4v3_encoder_select="h263_encoder" mss2_decoder_select="mpegvideo qpeldsp vc1_decoder" mts2_decoder_select="mss34dsp" +mvha_decoder_deps="zlib" mwsc_decoder_deps="zlib" mxpeg_decoder_select="mjpeg_decoder" nellymoser_decoder_select="mdct sinewin" diff --git a/libavcodec/Makefile b/libavcodec/Makefile index 52e5b4f345..c1f35b40d8 100644 --- a/libavcodec/Makefile +++ b/libavcodec/Makefile @@ -494,6 +494,7 @@ OBJS-$(CONFIG_MTS2_DECODER) += mss4.o OBJS-$(CONFIG_MVC1_DECODER) += mvcdec.o OBJS-$(CONFIG_MVC2_DECODER) += mvcdec.o OBJS-$(CONFIG_MVDV_DECODER) += midivid.o +OBJS-$(CONFIG_MVHA_DECODER) += mvha.o OBJS-$(CONFIG_MWSC_DECODER) += mwsc.o OBJS-$(CONFIG_MXPEG_DECODER) += mxpegdec.o OBJS-$(CONFIG_NELLYMOSER_DECODER) += nellymoserdec.o nellymoser.o diff --git a/libavcodec/allcodecs.c b/libavcodec/allcodecs.c index 4eb1afbea1..c33edf23c9 100644 --- a/libavcodec/allcodecs.c +++ b/libavcodec/allcodecs.c @@ -219,6 +219,7 @@ extern AVCodec ff_mts2_decoder; extern AVCodec ff_mvc1_decoder; extern AVCodec ff_mvc2_decoder; extern AVCodec ff_mvdv_decoder; +extern AVCodec ff_mvha_decoder; extern AVCodec ff_mwsc_decoder; extern AVCodec ff_mxpeg_decoder; extern AVCodec ff_nuv_decoder; diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h index 1cbc9c9ef1..735a3c2d76 100644 --- a/libavcodec/avcodec.h +++ b/libavcodec/avcodec.h @@ -459,6 +459,7 @@ enum AVCodecID { AV_CODEC_ID_VP4, AV_CODEC_ID_IMM5, AV_CODEC_ID_MVDV, + AV_CODEC_ID_MVHA, /* various PCM "codecs" */ AV_CODEC_ID_FIRST_AUDIO = 0x10000, ///< A dummy id pointing at the start of audio codecs diff --git a/libavcodec/codec_desc.c b/libavcodec/codec_desc.c index 3e634bbec7..570bd2f382 100644 --- a/libavcodec/codec_desc.c +++ b/libavcodec/codec_desc.c @@ -1740,6 +1740,13 @@ static const AVCodecDescriptor codec_descriptors[] = { .long_name = NULL_IF_CONFIG_SMALL("MidiVid VQ"), .props = AV_CODEC_PROP_LOSSY, }, + { + .id = AV_CODEC_ID_MVHA, + .type = AVMEDIA_TYPE_VIDEO, + .name = "mvha", + .long_name = NULL_IF_CONFIG_SMALL("MidiVid Archive Codec"), + .props = AV_CODEC_PROP_INTRA_ONLY | AV_CODEC_PROP_LOSSY, + }, /* various PCM "codecs" */ { diff --git a/libavcodec/mvha.c b/libavcodec/mvha.c new file mode 100644 index 0000000000..636b5e69d5 --- /dev/null +++ b/libavcodec/mvha.c @@ -0,0 +1,310 @@ +/* + * MidiVid Archive codec + * + * Copyright (c) 2019 Paul B Mahol + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#define CACHED_BITSTREAM_READER !ARCH_X86_32 +#include "libavutil/intreadwrite.h" + +#include "avcodec.h" +#include "bytestream.h" +#include "get_bits.h" +#include "internal.h" +#include "lossless_videodsp.h" + +#include <zlib.h> + +typedef struct MVHAContext { + GetBitContext gb; + int nb_symbols; + + uint8_t symb[256]; + uint32_t prob[256]; + VLC vlc; + + z_stream zstream; + LLVidDSPContext llviddsp; +} MVHAContext; + +typedef struct Node { + int16_t sym; + int16_t n0; + int16_t l, r; + uint32_t count; +} Node; + +static void get_tree_codes(uint32_t *bits, int16_t *lens, uint8_t *xlat, + Node *nodes, int node, + uint32_t pfx, int pl, int *pos) +{ + int s; + + s = nodes[node].sym; + if (s != -1) { + bits[*pos] = (~pfx) & ((1ULL << FFMAX(pl, 1)) - 1); + lens[*pos] = FFMAX(pl, 1); + xlat[*pos] = s + (pl == 0); + (*pos)++; + } else { + pfx <<= 1; + pl++; + get_tree_codes(bits, lens, xlat, nodes, nodes[node].l, pfx, pl, + pos); + pfx |= 1; + get_tree_codes(bits, lens, xlat, nodes, nodes[node].r, pfx, pl, + pos); + } +} + +static int build_vlc(AVCodecContext *avctx, VLC *vlc) +{ + MVHAContext *s = avctx->priv_data; + Node nodes[512]; + uint32_t bits[256]; + int16_t lens[256]; + uint8_t xlat[256]; + int cur_node, i, j, pos = 0; + + ff_free_vlc(vlc); + + for (i = 0; i < s->nb_symbols; i++) { + nodes[i].count = s->prob[i]; + nodes[i].sym = s->symb[i]; + nodes[i].n0 = -2; + nodes[i].l = i; + nodes[i].r = i; + } + + cur_node = s->nb_symbols; + j = 0; + do { + for (i = 0; ; i++) { + int new_node = j; + int first_node = cur_node; + int second_node = cur_node; + unsigned nd, st; + + nodes[cur_node].count = -1; + + do { + int val = nodes[new_node].count; + if (val && (val < nodes[first_node].count)) { + if (val >= nodes[second_node].count) { + first_node = new_node; + } else { + first_node = second_node; + second_node = new_node; + } + } + new_node += 1; + } while (new_node != cur_node); + + if (first_node == cur_node) + break; + + nd = nodes[second_node].count; + st = nodes[first_node].count; + nodes[second_node].count = 0; + nodes[first_node].count = 0; + if (nd >= UINT32_MAX - st) { + av_log(avctx, AV_LOG_ERROR, "count overflow\n"); + return AVERROR_INVALIDDATA; + } + nodes[cur_node].count = nd + st; + nodes[cur_node].sym = -1; + nodes[cur_node].n0 = cur_node; + nodes[cur_node].l = first_node; + nodes[cur_node].r = second_node; + cur_node++; + } + j++; + } while (cur_node - s->nb_symbols == j); + + get_tree_codes(bits, lens, xlat, nodes, cur_node - 1, 0, 0, &pos); + + return ff_init_vlc_sparse(vlc, 12, pos, lens, 2, 2, bits, 4, 4, xlat, 1, 1, 0); +} + +static int decode_frame(AVCodecContext *avctx, + void *data, int *got_frame, + AVPacket *avpkt) +{ + MVHAContext *s = avctx->priv_data; + AVFrame *frame = data; + uint32_t type, size; + int ret; + + if (avpkt->size <= 8) + return AVERROR_INVALIDDATA; + + type = AV_RB32(avpkt->data); + size = AV_RL32(avpkt->data + 4); + + if ((ret = ff_get_buffer(avctx, frame, 0)) < 0) + return ret; + + if (type == MKTAG('L','Z','Y','V')) { + ret = inflateReset(&s->zstream); + if (ret != Z_OK) { + av_log(avctx, AV_LOG_ERROR, "Inflate reset error: %d\n", ret); + return AVERROR_EXTERNAL; + } + + s->zstream.next_in = avpkt->data + 8; + s->zstream.avail_in = avpkt->size - 8; + + for (int p = 0; p < 3; p++) { + for (int y = 0; y < avctx->height; y++) { + s->zstream.next_out = frame->data[p] + (avctx->height - y - 1) * frame->linesize[p]; + s->zstream.avail_out = avctx->width >> (p > 0); + + ret = inflate(&s->zstream, Z_SYNC_FLUSH); + if (ret != Z_OK && ret != Z_STREAM_END) { + av_log(avctx, AV_LOG_ERROR, "Inflate error: %d\n", ret); + return AVERROR_EXTERNAL; + } + } + } + } else if (type == MKTAG('H','U','F','Y')) { + GetBitContext *gb = &s->gb; + int first_symbol, symbol; + + init_get_bits8(gb, avpkt->data + 8, avpkt->size - 8); + + skip_bits(gb, 24); + + first_symbol = get_bits(gb, 8); + s->nb_symbols = get_bits(gb, 8) + 1; + + symbol = first_symbol; + for (int i = 0; i < s->nb_symbols; symbol++) { + int prob; + + if (get_bits1(gb)) { + prob = get_bits(gb, 12); + } else { + prob = get_bits(gb, 3); + } + + if (prob) { + s->symb[i] = symbol; + s->prob[i] = prob; + i++; + } + } + + ret = build_vlc(avctx, &s->vlc); + if (ret < 0) + return ret; + + for (int p = 0; p < 3; p++) { + int width = avctx->width >> (p > 0); + ptrdiff_t stride = frame->linesize[p]; + uint8_t *dst; + + dst = frame->data[p] + (avctx->height - 1) * frame->linesize[p]; + for (int y = 0; y < avctx->height; y++) { + for (int x = 0; x < width; x++) { + int v = get_vlc2(gb, s->vlc.table, s->vlc.bits, 3); + + if (v < 0) + return AVERROR_INVALIDDATA; + + dst[x] = v; + } + dst -= stride; + } + } + } else { + return AVERROR_INVALIDDATA; + } + + for (int p = 0; p < 3; p++) { + int left, lefttop; + int width = avctx->width >> (p > 0); + ptrdiff_t stride = frame->linesize[p]; + uint8_t *dst; + + dst = frame->data[p] + (avctx->height - 1) * frame->linesize[p]; + s->llviddsp.add_left_pred(dst, dst, width, 0); + dst -= stride; + lefttop = left = dst[0]; + for (int y = 1; y < avctx->height; y++) { + s->llviddsp.add_median_pred(dst, dst + stride, dst, width, &left, &lefttop); + lefttop = left = dst[0]; + dst -= stride; + } + } + + frame->pict_type = AV_PICTURE_TYPE_I; + frame->key_frame = 1; + *got_frame = 1; + + return avpkt->size; +} + +static av_cold int decode_init(AVCodecContext *avctx) +{ + MVHAContext *s = avctx->priv_data; + int zret; + + avctx->pix_fmt = AV_PIX_FMT_YUV422P; + + s->zstream.zalloc = Z_NULL; + s->zstream.zfree = Z_NULL; + s->zstream.opaque = Z_NULL; + zret = inflateInit(&s->zstream); + if (zret != Z_OK) { + av_log(avctx, AV_LOG_ERROR, "Inflate init error: %d\n", zret); + return AVERROR_EXTERNAL; + } + + ff_llviddsp_init(&s->llviddsp); + + return 0; +} + +static av_cold int decode_close(AVCodecContext *avctx) +{ + MVHAContext *s = avctx->priv_data; + + inflateEnd(&s->zstream); + ff_free_vlc(&s->vlc); + + return 0; +} + +AVCodec ff_mvha_decoder = { + .name = "mvha", + .long_name = NULL_IF_CONFIG_SMALL("MidiVid Archive Codec"), + .type = AVMEDIA_TYPE_VIDEO, + .id = AV_CODEC_ID_MVHA, + .priv_data_size = sizeof(MVHAContext), + .init = decode_init, + .close = decode_close, + .decode = decode_frame, + .capabilities = AV_CODEC_CAP_DR1, + .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE | + FF_CODEC_CAP_INIT_CLEANUP, +}; diff --git a/libavcodec/version.h b/libavcodec/version.h index 5c8e376ea1..efa9d743f7 100644 --- a/libavcodec/version.h +++ b/libavcodec/version.h @@ -28,7 +28,7 @@ #include "libavutil/version.h" #define LIBAVCODEC_VERSION_MAJOR 58 -#define LIBAVCODEC_VERSION_MINOR 63 +#define LIBAVCODEC_VERSION_MINOR 64 #define LIBAVCODEC_VERSION_MICRO 100 #define LIBAVCODEC_VERSION_INT AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \ diff --git a/libavformat/riff.c b/libavformat/riff.c index 25ccedc8ce..c73f6e9db0 100644 --- a/libavformat/riff.c +++ b/libavformat/riff.c @@ -490,6 +490,7 @@ const AVCodecTag ff_codec_bmp_tags[] = { { AV_CODEC_ID_LSCR, MKTAG('L', 'S', 'C', 'R') }, { AV_CODEC_ID_IMM5, MKTAG('I', 'M', 'M', '5') }, { AV_CODEC_ID_MVDV, MKTAG('M', 'V', 'D', 'V') }, + { AV_CODEC_ID_MVHA, MKTAG('M', 'V', 'H', 'A') }, { AV_CODEC_ID_NONE, 0 } };

1 0

avcodec: add mvdv video decoder
by Paul B Mahol 27 Nov '19

27 Nov '19

ffmpeg | branch: master | Paul B Mahol <onemda(a)gmail.com> | Sat Nov 23 13:46:55 2019 +0100| [61dc7add30208a2ce24e0d1d14524e80832ba1b1] | committer: Paul B Mahol avcodec: add mvdv video decoder > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=61dc7add30208a2ce… --- Changelog | 1 + libavcodec/Makefile | 1 + libavcodec/allcodecs.c | 1 + libavcodec/avcodec.h | 1 + libavcodec/codec_desc.c | 7 ++ libavcodec/midivid.c | 283 ++++++++++++++++++++++++++++++++++++++++++++++++ libavcodec/version.h | 2 +- libavformat/riff.c | 1 + 8 files changed, 296 insertions(+), 1 deletion(-) diff --git a/Changelog b/Changelog index 1e9476b24d..91d1b5a51e 100644 --- a/Changelog +++ b/Changelog @@ -25,6 +25,7 @@ version <next>: - AV1 frame merge bitstream filter - AV1 Annex B demuxer - axcorrelate filter +- mvdv decoder version 4.2: diff --git a/libavcodec/Makefile b/libavcodec/Makefile index 006a472a6d..52e5b4f345 100644 --- a/libavcodec/Makefile +++ b/libavcodec/Makefile @@ -493,6 +493,7 @@ OBJS-$(CONFIG_MSZH_DECODER) += lcldec.o OBJS-$(CONFIG_MTS2_DECODER) += mss4.o OBJS-$(CONFIG_MVC1_DECODER) += mvcdec.o OBJS-$(CONFIG_MVC2_DECODER) += mvcdec.o +OBJS-$(CONFIG_MVDV_DECODER) += midivid.o OBJS-$(CONFIG_MWSC_DECODER) += mwsc.o OBJS-$(CONFIG_MXPEG_DECODER) += mxpegdec.o OBJS-$(CONFIG_NELLYMOSER_DECODER) += nellymoserdec.o nellymoser.o diff --git a/libavcodec/allcodecs.c b/libavcodec/allcodecs.c index 0c0741936c..4eb1afbea1 100644 --- a/libavcodec/allcodecs.c +++ b/libavcodec/allcodecs.c @@ -218,6 +218,7 @@ extern AVCodec ff_mszh_decoder; extern AVCodec ff_mts2_decoder; extern AVCodec ff_mvc1_decoder; extern AVCodec ff_mvc2_decoder; +extern AVCodec ff_mvdv_decoder; extern AVCodec ff_mwsc_decoder; extern AVCodec ff_mxpeg_decoder; extern AVCodec ff_nuv_decoder; diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h index 813a43b72e..1cbc9c9ef1 100644 --- a/libavcodec/avcodec.h +++ b/libavcodec/avcodec.h @@ -458,6 +458,7 @@ enum AVCodecID { AV_CODEC_ID_LSCR, AV_CODEC_ID_VP4, AV_CODEC_ID_IMM5, + AV_CODEC_ID_MVDV, /* various PCM "codecs" */ AV_CODEC_ID_FIRST_AUDIO = 0x10000, ///< A dummy id pointing at the start of audio codecs diff --git a/libavcodec/codec_desc.c b/libavcodec/codec_desc.c index 5961af3c85..3e634bbec7 100644 --- a/libavcodec/codec_desc.c +++ b/libavcodec/codec_desc.c @@ -1733,6 +1733,13 @@ static const AVCodecDescriptor codec_descriptors[] = { .long_name = NULL_IF_CONFIG_SMALL("Infinity IMM5"), .props = AV_CODEC_PROP_LOSSY, }, + { + .id = AV_CODEC_ID_MVDV, + .type = AVMEDIA_TYPE_VIDEO, + .name = "mvdv", + .long_name = NULL_IF_CONFIG_SMALL("MidiVid VQ"), + .props = AV_CODEC_PROP_LOSSY, + }, /* various PCM "codecs" */ { diff --git a/libavcodec/midivid.c b/libavcodec/midivid.c new file mode 100644 index 0000000000..38465c5fbc --- /dev/null +++ b/libavcodec/midivid.c @@ -0,0 +1,283 @@ +/* + * MidiVid decoder + * Copyright (c) 2019 Paul B Mahol + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#include "libavutil/imgutils.h" +#include "libavutil/internal.h" +#include "libavutil/intreadwrite.h" +#include "libavutil/mem.h" + +#define BITSTREAM_READER_LE +#include "avcodec.h" +#include "get_bits.h" +#include "bytestream.h" +#include "internal.h" + +typedef struct MidiVidContext { + GetByteContext gb; + + uint8_t *uncompressed; + unsigned int uncompressed_size; + uint8_t *skip; + + AVFrame *frame; +} MidiVidContext; + +static int decode_mvdv(MidiVidContext *s, AVCodecContext *avctx, AVFrame *frame) +{ + GetByteContext *gb = &s->gb; + GetBitContext mask; + GetByteContext idx9; + uint16_t nb_vectors, intra_flag; + const uint8_t *vec; + const uint8_t *mask_start; + uint8_t *skip; + uint32_t mask_size; + int idx9bits = 0; + int idx9val = 0; + uint32_t nb_blocks; + + nb_vectors = bytestream2_get_le16(gb); + intra_flag = bytestream2_get_le16(gb); + if (intra_flag) { + nb_blocks = (avctx->width / 2) * (avctx->height / 2); + } else { + int skip_linesize; + + nb_blocks = bytestream2_get_le32(gb); + skip_linesize = avctx->width >> 1; + mask_start = gb->buffer_start + bytestream2_tell(gb); + mask_size = (avctx->width >> 5) * (avctx->height >> 2); + + if (bytestream2_get_bytes_left(gb) < mask_size) + return AVERROR_INVALIDDATA; + + init_get_bits8(&mask, mask_start, mask_size); + bytestream2_skip(gb, mask_size); + skip = s->skip; + + for (int y = 0; y < avctx->height >> 2; y++) { + for (int x = 0; x < avctx->width >> 2; x++) { + int flag = !get_bits1(&mask); + + skip[(y*2) *skip_linesize + x*2 ] = flag; + skip[(y*2) *skip_linesize + x*2+1] = flag; + skip[(y*2+1)*skip_linesize + x*2 ] = flag; + skip[(y*2+1)*skip_linesize + x*2+1] = flag; + } + } + } + + vec = gb->buffer_start + bytestream2_tell(gb); + if (bytestream2_get_bytes_left(gb) < nb_vectors * 12) + return AVERROR_INVALIDDATA; + bytestream2_skip(gb, nb_vectors * 12); + if (nb_vectors > 256) { + if (bytestream2_get_bytes_left(gb) < (nb_blocks + 7) / 8) + return AVERROR_INVALIDDATA; + bytestream2_init(&idx9, gb->buffer_start + bytestream2_tell(gb), (nb_blocks + 7) / 8); + bytestream2_skip(gb, (nb_blocks + 7) / 8); + } + + skip = s->skip; + + for (int y = avctx->height - 2; y >= 0; y -= 2) { + uint8_t *dsty = frame->data[0] + y * frame->linesize[0]; + uint8_t *dstu = frame->data[1] + y * frame->linesize[1]; + uint8_t *dstv = frame->data[2] + y * frame->linesize[2]; + + for (int x = 0; x < avctx->width; x += 2) { + int idx; + + if (!intra_flag && *skip++) + continue; + if (bytestream2_get_bytes_left(gb) <= 0) + return AVERROR_INVALIDDATA; + if (nb_vectors <= 256) { + idx = bytestream2_get_byte(gb); + } else { + if (idx9bits == 0) { + idx9val = bytestream2_get_byte(&idx9); + idx9bits = 8; + } + idx9bits--; + idx = bytestream2_get_byte(gb) | (((idx9val >> (7 - idx9bits)) & 1) << 8); + } + + dsty[x +frame->linesize[0]] = vec[idx * 12 + 0]; + dsty[x+1+frame->linesize[0]] = vec[idx * 12 + 3]; + dsty[x] = vec[idx * 12 + 6]; + dsty[x+1] = vec[idx * 12 + 9]; + + dstu[x +frame->linesize[1]] = vec[idx * 12 + 1]; + dstu[x+1+frame->linesize[1]] = vec[idx * 12 + 4]; + dstu[x] = vec[idx * 12 + 7]; + dstu[x+1] = vec[idx * 12 +10]; + + dstv[x +frame->linesize[2]] = vec[idx * 12 + 2]; + dstv[x+1+frame->linesize[2]] = vec[idx * 12 + 5]; + dstv[x] = vec[idx * 12 + 8]; + dstv[x+1] = vec[idx * 12 +11]; + } + } + + return intra_flag; +} + +static ptrdiff_t lzss_uncompress(MidiVidContext *s, GetByteContext *gb, uint8_t *dst, unsigned int size) +{ + uint8_t *dst_start = dst; + uint8_t *dst_end = dst + size; + + for (;bytestream2_get_bytes_left(gb) >= 3;) { + int op = bytestream2_get_le16(gb); + + for (int i = 0; i < 16; i++) { + if (op & 1) { + int s0 = bytestream2_get_byte(gb); + int s1 = bytestream2_get_byte(gb); + int offset = ((s0 & 0xF0) << 4) | s1; + int length = (s0 & 0xF) + 3; + + if (dst + length > dst_end || + dst - offset < dst_start) + return AVERROR_INVALIDDATA; + if (offset > 0) { + for (int j = 0; j < length; j++) { + dst[j] = dst[j - offset]; + } + } + dst += length; + } else { + if (dst >= dst_end) + return AVERROR_INVALIDDATA; + *dst++ = bytestream2_get_byte(gb); + } + op >>= 1; + } + } + + return dst - dst_start; +} + +static int decode_frame(AVCodecContext *avctx, void *data, + int *got_frame, AVPacket *avpkt) +{ + MidiVidContext *s = avctx->priv_data; + GetByteContext *gb = &s->gb; + AVFrame *frame = s->frame; + int ret, key, uncompressed; + + if (avpkt->size <= 13) + return AVERROR_INVALIDDATA; + + bytestream2_init(gb, avpkt->data, avpkt->size); + bytestream2_skip(gb, 8); + uncompressed = bytestream2_get_le32(gb); + + if ((ret = ff_reget_buffer(avctx, s->frame, 0)) < 0) + return ret; + + if (uncompressed) { + ret = decode_mvdv(s, avctx, frame); + } else { + av_fast_padded_malloc(&s->uncompressed, &s->uncompressed_size, 16LL * (avpkt->size - 12)); + if (!s->uncompressed) + return AVERROR(ENOMEM); + + ret = lzss_uncompress(s, gb, s->uncompressed, s->uncompressed_size); + if (ret < 0) + return ret; + bytestream2_init(gb, s->uncompressed, ret); + ret = decode_mvdv(s, avctx, frame); + } + + if (ret < 0) + return ret; + key = ret; + + if ((ret = av_frame_ref(data, s->frame)) < 0) + return ret; + + frame->pict_type = key ? AV_PICTURE_TYPE_I : AV_PICTURE_TYPE_P; + frame->key_frame = key; + *got_frame = 1; + + return avpkt->size; +} + +static av_cold int decode_init(AVCodecContext *avctx) +{ + MidiVidContext *s = avctx->priv_data; + int ret = av_image_check_size(avctx->width, avctx->height, 0, avctx); + + if (ret < 0) { + av_log(avctx, AV_LOG_ERROR, "Invalid image size %dx%d.\n", + avctx->width, avctx->height); + return ret; + } + + avctx->pix_fmt = AV_PIX_FMT_YUV444P; + + s->frame = av_frame_alloc(); + if (!s->frame) + return AVERROR(ENOMEM); + s->skip = av_calloc(avctx->width >> 1, avctx->height >> 1); + if (!s->skip) + return AVERROR(ENOMEM); + + return 0; +} + +static void decode_flush(AVCodecContext *avctx) +{ + MidiVidContext *s = avctx->priv_data; + + av_frame_unref(s->frame); +} + +static av_cold int decode_close(AVCodecContext *avctx) +{ + MidiVidContext *s = avctx->priv_data; + + av_frame_free(&s->frame); + av_freep(&s->uncompressed); + av_freep(&s->skip); + + return 0; +} + +AVCodec ff_mvdv_decoder = { + .name = "mvdv", + .long_name = NULL_IF_CONFIG_SMALL("MidiVid VQ"), + .type = AVMEDIA_TYPE_VIDEO, + .id = AV_CODEC_ID_MVDV, + .priv_data_size = sizeof(MidiVidContext), + .init = decode_init, + .decode = decode_frame, + .flush = decode_flush, + .close = decode_close, + .capabilities = AV_CODEC_CAP_DR1, + .caps_internal = FF_CODEC_CAP_INIT_CLEANUP, +}; diff --git a/libavcodec/version.h b/libavcodec/version.h index 58ea00a520..5c8e376ea1 100644 --- a/libavcodec/version.h +++ b/libavcodec/version.h @@ -28,7 +28,7 @@ #include "libavutil/version.h" #define LIBAVCODEC_VERSION_MAJOR 58 -#define LIBAVCODEC_VERSION_MINOR 62 +#define LIBAVCODEC_VERSION_MINOR 63 #define LIBAVCODEC_VERSION_MICRO 100 #define LIBAVCODEC_VERSION_INT AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \ diff --git a/libavformat/riff.c b/libavformat/riff.c index 048a79e394..25ccedc8ce 100644 --- a/libavformat/riff.c +++ b/libavformat/riff.c @@ -489,6 +489,7 @@ const AVCodecTag ff_codec_bmp_tags[] = { { AV_CODEC_ID_AGM, MKTAG('A', 'G', 'M', '7') }, { AV_CODEC_ID_LSCR, MKTAG('L', 'S', 'C', 'R') }, { AV_CODEC_ID_IMM5, MKTAG('I', 'M', 'M', '5') }, + { AV_CODEC_ID_MVDV, MKTAG('M', 'V', 'D', 'V') }, { AV_CODEC_ID_NONE, 0 } };

1 0

avformat/hls: correct grammatical errors of m3u8_hold_counters option
by Steven Liu 27 Nov '19

27 Nov '19

ffmpeg | branch: master | Steven Liu <lq(a)chinaffmpeg.org> | Wed Nov 27 19:04:00 2019 +0800| [afee801276d5aea66057aa9d4d37a13cf7ae7372] | committer: Steven Liu avformat/hls: correct grammatical errors of m3u8_hold_counters option Suggested-by: Gyan <ffmpeg(a)gyani.pro> Suggested-by: Rodney Baker <rodney.baker(a)iinet.net.au> Signed-off-by: Steven Liu <lq(a)chinaffmpeg.org> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=afee801276d5aea66… --- doc/demuxers.texi | 2 +- libavformat/hls.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/doc/demuxers.texi b/doc/demuxers.texi index 4e1a5cb6aa..0d13bdd1b3 100644 --- a/doc/demuxers.texi +++ b/doc/demuxers.texi @@ -332,7 +332,7 @@ Maximum number of times a insufficient list is attempted to be reloaded. Default value is 1000. @item m3u8_hold_counters -Maximum number of times to load m3u8 when the m3u8 dose not refresh with new segments. +The maximum number of times to load m3u8 when it refreshes without new segments. Default value is 1000. @item http_persistent diff --git a/libavformat/hls.c b/libavformat/hls.c index 0b1582b2b5..21353bbad7 100644 --- a/libavformat/hls.c +++ b/libavformat/hls.c @@ -2332,7 +2332,7 @@ static const AVOption hls_options[] = { INT_MIN, INT_MAX, FLAGS}, {"max_reload", "Maximum number of times a insufficient list is attempted to be reloaded", OFFSET(max_reload), AV_OPT_TYPE_INT, {.i64 = 1000}, 0, INT_MAX, FLAGS}, - {"m3u8_hold_counters", "Maximum number of times requests when the m3u8 file not be refresh", + {"m3u8_hold_counters", "The maximum number of times to load m3u8 when it refreshes without new segments", OFFSET(m3u8_hold_counters), AV_OPT_TYPE_INT, {.i64 = 1000}, 0, INT_MAX, FLAGS}, {"http_persistent", "Use persistent HTTP connections", OFFSET(http_persistent), AV_OPT_TYPE_BOOL, {.i64 = 1}, 0, 1, FLAGS },

1 0

lavc/qsvenc: Fix compilation for some build environments.
by Carl Eugen Hoyos 26 Nov '19

26 Nov '19

ffmpeg | branch: master | Carl Eugen Hoyos <ceffmpeg(a)gmail.com> | Wed Nov 27 00:03:33 2019 +0100| [a2fbdc6898fc72f7f289e750f2199ee66216f2f8] | committer: Carl Eugen Hoyos lavc/qsvenc: Fix compilation for some build environments. Reported and tested by Sean Darcy. > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=a2fbdc6898fc72f7f… --- libavcodec/qsvenc.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/libavcodec/qsvenc.c b/libavcodec/qsvenc.c index 93d49ba21d..1bc5600061 100644 --- a/libavcodec/qsvenc.c +++ b/libavcodec/qsvenc.c @@ -824,7 +824,9 @@ static int qsv_retrieve_enc_vp9_params(AVCodecContext *avctx, QSVEncContext *q) #endif mfxExtBuffer *ext_buffers[] = { +#if QSV_HAVE_EXT_VP9_PARAM (mfxExtBuffer*)&vp9_extend_buf, +#endif #if QSV_HAVE_CO2 (mfxExtBuffer*)&co2, #endif

1 0

avutil/hwcontext_cuda: allow using primary CUDA device context
by Oleg Dobkin 26 Nov '19

26 Nov '19

ffmpeg | branch: master | Oleg Dobkin <olegd(a)anyvision.co> | Mon Nov 18 15:35:49 2019 +0200| [32ba563cfccd41cc9b8c69674495fc1d8d705283] | committer: Timo Rothenpieler avutil/hwcontext_cuda: allow using primary CUDA device context Signed-off-by: Timo Rothenpieler <timo(a)rothenpieler.org> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=32ba563cfccd41cc9… --- configure | 8 +++--- libavutil/hwcontext_cuda.c | 49 +++++++++++++++++++++++++++++-------- libavutil/hwcontext_cuda.h | 17 +++++++++++++ libavutil/hwcontext_cuda_internal.h | 2 ++ libavutil/version.h | 2 +- 5 files changed, 63 insertions(+), 15 deletions(-) diff --git a/configure b/configure index 8f4f2884cf..2519e6421f 100755 --- a/configure +++ b/configure @@ -6139,10 +6139,10 @@ fi if ! disabled ffnvcodec; then ffnv_hdr_list="ffnvcodec/nvEncodeAPI.h ffnvcodec/dynlink_cuda.h ffnvcodec/dynlink_cuviddec.h ffnvcodec/dynlink_nvcuvid.h" - check_pkg_config ffnvcodec "ffnvcodec >= 9.0.18.0" "$ffnv_hdr_list" "" || \ - check_pkg_config ffnvcodec "ffnvcodec >= 8.2.15.8 ffnvcodec < 8.3" "$ffnv_hdr_list" "" || \ - check_pkg_config ffnvcodec "ffnvcodec >= 8.1.24.9 ffnvcodec < 8.2" "$ffnv_hdr_list" "" || \ - check_pkg_config ffnvcodec "ffnvcodec >= 8.0.14.9 ffnvcodec < 8.1" "$ffnv_hdr_list" "" + check_pkg_config ffnvcodec "ffnvcodec >= 9.1.23.1" "$ffnv_hdr_list" "" || \ + check_pkg_config ffnvcodec "ffnvcodec >= 9.0.18.3 ffnvcodec < 9.1" "$ffnv_hdr_list" "" || \ + check_pkg_config ffnvcodec "ffnvcodec >= 8.2.15.10 ffnvcodec < 8.3" "$ffnv_hdr_list" "" || \ + check_pkg_config ffnvcodec "ffnvcodec >= 8.1.24.11 ffnvcodec < 8.2" "$ffnv_hdr_list" "" fi check_cpp_condition winrt windows.h "!WINAPI_FAMILY_PARTITION(WINAPI_PARTITION_DESKTOP)" diff --git a/libavutil/hwcontext_cuda.c b/libavutil/hwcontext_cuda.c index cca39e9fc7..30611b1912 100644 --- a/libavutil/hwcontext_cuda.c +++ b/libavutil/hwcontext_cuda.c @@ -280,10 +280,16 @@ static void cuda_device_uninit(AVHWDeviceContext *device_ctx) if (hwctx->internal) { CudaFunctions *cu = hwctx->internal->cuda_dl; + if (hwctx->internal->is_allocated && hwctx->cuda_ctx) { - CHECK_CU(cu->cuCtxDestroy(hwctx->cuda_ctx)); + if (hwctx->internal->flags & AV_CUDA_USE_PRIMARY_CONTEXT) + CHECK_CU(cu->cuDevicePrimaryCtxRelease(hwctx->internal->cuda_device)); + else + CHECK_CU(cu->cuCtxDestroy(hwctx->cuda_ctx)); + hwctx->cuda_ctx = NULL; } + cuda_free_functions(&hwctx->internal->cuda_dl); } @@ -322,9 +328,11 @@ static int cuda_device_create(AVHWDeviceContext *device_ctx, { AVCUDADeviceContext *hwctx = device_ctx->hwctx; CudaFunctions *cu; - CUdevice cu_device; CUcontext dummy; - int ret, device_idx = 0; + int ret, dev_active = 0, device_idx = 0; + unsigned int dev_flags = 0; + + const unsigned int desired_flags = CU_CTX_SCHED_BLOCKING_SYNC; if (device) device_idx = strtol(device, NULL, 0); @@ -338,21 +346,42 @@ static int cuda_device_create(AVHWDeviceContext *device_ctx, if (ret < 0) goto error; - ret = CHECK_CU(cu->cuDeviceGet(&cu_device, device_idx)); + ret = CHECK_CU(cu->cuDeviceGet(&hwctx->internal->cuda_device, device_idx)); if (ret < 0) goto error; - ret = CHECK_CU(cu->cuCtxCreate(&hwctx->cuda_ctx, CU_CTX_SCHED_BLOCKING_SYNC, cu_device)); - if (ret < 0) - goto error; + hwctx->internal->flags = flags; - // Setting stream to NULL will make functions automatically use the default CUstream - hwctx->stream = NULL; + if (flags & AV_CUDA_USE_PRIMARY_CONTEXT) { + ret = CHECK_CU(cu->cuDevicePrimaryCtxGetState(hwctx->internal->cuda_device, &dev_flags, &dev_active)); + if (ret < 0) + goto error; - CHECK_CU(cu->cuCtxPopCurrent(&dummy)); + if (dev_active && dev_flags != desired_flags) { + av_log(device_ctx, AV_LOG_ERROR, "Primary context already active with incompatible flags.\n"); + goto error; + } else if (dev_flags != desired_flags) { + ret = CHECK_CU(cu->cuDevicePrimaryCtxSetFlags(hwctx->internal->cuda_device, desired_flags)); + if (ret < 0) + goto error; + } + + ret = CHECK_CU(cu->cuDevicePrimaryCtxRetain(&hwctx->cuda_ctx, hwctx->internal->cuda_device)); + if (ret < 0) + goto error; + } else { + ret = CHECK_CU(cu->cuCtxCreate(&hwctx->cuda_ctx, desired_flags, hwctx->internal->cuda_device)); + if (ret < 0) + goto error; + + CHECK_CU(cu->cuCtxPopCurrent(&dummy)); + } hwctx->internal->is_allocated = 1; + // Setting stream to NULL will make functions automatically use the default CUstream + hwctx->stream = NULL; + return 0; error: diff --git a/libavutil/hwcontext_cuda.h b/libavutil/hwcontext_cuda.h index 81a0552cab..cefbe0ceab 100644 --- a/libavutil/hwcontext_cuda.h +++ b/libavutil/hwcontext_cuda.h @@ -49,4 +49,21 @@ typedef struct AVCUDADeviceContext { * AVHWFramesContext.hwctx is currently not used */ +/** + * @defgroup hwcontext_cuda Device context creation flags + * + * Flags for av_hwdevice_ctx_create. + * + * @{ + */ + +/** + * Use primary device context instead of creating a new one. + */ +#define AV_CUDA_USE_PRIMARY_CONTEXT (1 << 0) + +/** + * @} + */ + #endif /* AVUTIL_HWCONTEXT_CUDA_H */ diff --git a/libavutil/hwcontext_cuda_internal.h b/libavutil/hwcontext_cuda_internal.h index e1bc6ff350..d5633c58d5 100644 --- a/libavutil/hwcontext_cuda_internal.h +++ b/libavutil/hwcontext_cuda_internal.h @@ -31,6 +31,8 @@ struct AVCUDADeviceContextInternal { CudaFunctions *cuda_dl; int is_allocated; + CUdevice cuda_device; + int flags; }; #endif /* AVUTIL_HWCONTEXT_CUDA_INTERNAL_H */ diff --git a/libavutil/version.h b/libavutil/version.h index af3abf7265..e18163388d 100644 --- a/libavutil/version.h +++ b/libavutil/version.h @@ -80,7 +80,7 @@ #define LIBAVUTIL_VERSION_MAJOR 56 #define LIBAVUTIL_VERSION_MINOR 36 -#define LIBAVUTIL_VERSION_MICRO 100 +#define LIBAVUTIL_VERSION_MICRO 101 #define LIBAVUTIL_VERSION_INT AV_VERSION_INT(LIBAVUTIL_VERSION_MAJOR, \ LIBAVUTIL_VERSION_MINOR, \

1 0

avformat/hlsenc: set strict_std_compliance from the parent AVFormatContext
by Steven Liu 26 Nov '19

26 Nov '19

ffmpeg | branch: master | Steven Liu <lq(a)chinaffmpeg.org> | Mon Nov 25 11:15:13 2019 +0800| [59d264b0a1218094709d61e30c9056c90859567f] | committer: Steven Liu avformat/hlsenc: set strict_std_compliance from the parent AVFormatContext fix ticket: 8388 Reviewed-by: Jan Ekström <jeebjp(a)gmail.com> Signed-off-by: Steven Liu <lq(a)chinaffmpeg.org> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=59d264b0a12180947… --- libavformat/hlsenc.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/libavformat/hlsenc.c b/libavformat/hlsenc.c index 013e91ad88..043438368e 100644 --- a/libavformat/hlsenc.c +++ b/libavformat/hlsenc.c @@ -776,12 +776,13 @@ static int hls_mux_init(AVFormatContext *s, VariantStream *vs) if (!oc->url) return AVERROR(ENOMEM); - oc->oformat = vs->oformat; - oc->interrupt_callback = s->interrupt_callback; - oc->max_delay = s->max_delay; - oc->opaque = s->opaque; - oc->io_open = s->io_open; - oc->io_close = s->io_close; + oc->oformat = vs->oformat; + oc->interrupt_callback = s->interrupt_callback; + oc->max_delay = s->max_delay; + oc->opaque = s->opaque; + oc->io_open = s->io_open; + oc->io_close = s->io_close; + oc->strict_std_compliance = s->strict_std_compliance; av_dict_copy(&oc->metadata, s->metadata, 0); if (vs->vtt_oformat) {

1 0

avcodec/v210dec: add support for frame and slice threading
by Limin Wang 25 Nov '19

25 Nov '19

ffmpeg | branch: master | Limin Wang <lance.lmwang(a)gmail.com> | Mon Nov 25 11:45:10 2019 +0800| [d07d1c1dd06541b3b48f3eacb7d8747a3456c6ad] | committer: Michael Niedermayer avcodec/v210dec: add support for frame and slice threading 1, Test server configure: [root@localhost ~]# cat /proc/cpuinfo |grep "model name" model name : Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz model name : Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz ... [root@localhost ~]# free -h total used free shared buff/cache available Mem: 102G 1.1G 100G 16M 657M 100G Swap: 4.0G 0B 4.0G 2, Test result: encode the v210 input data for testing: ./ffmpeg -y -i 4k_422.ts -c:v v210 -vframes 10 test.avi master: ./ffmpeg -y -threads 1 -stream_loop 1000 -i ./test.avi -benchmark -f null - frame=10010 fps= 60 q=-0.0 Lsize=N/A time=00:38:26.30 bitrate=N/A speed=13.7x video:5240kB audio:432432kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown bench: utime=101.869s stime=66.181s rtime=167.996s bench: maxrss=186552kB patch applied: ./ffmpeg -y -threads 2 -thread_type slice -stream_loop 1000 -i ./test.avi -benchmark -f null - frame=10010 fps= 72 q=-0.0 Lsize=N/A time=00:38:26.30 bitrate=N/A speed=16.5x video:5240kB audio:432432kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown bench: utime=103.562s stime=74.858s rtime=139.599s bench: maxrss=188616kB ./ffmpeg -y -threads 2 -thread_type frame -stream_loop 1000 -i ./test.avi -benchmark -f null - frame=10010 fps= 85 q=-0.0 Lsize=N/A time=00:38:26.30 bitrate=N/A speed=19.6x video:5240kB audio:432432kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown bench: utime=114.310s stime=92.685s rtime=117.693s bench: maxrss=231896kB Signed-off-by: Limin Wang <lance.lmwang(a)gmail.com> Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=d07d1c1dd06541b3b… --- libavcodec/v210dec.c | 126 +++++++++++++++++++++++++++++++-------------------- libavcodec/v210dec.h | 1 + 2 files changed, 79 insertions(+), 48 deletions(-) diff --git a/libavcodec/v210dec.c b/libavcodec/v210dec.c index 5a33d8c089..8483023815 100644 --- a/libavcodec/v210dec.c +++ b/libavcodec/v210dec.c @@ -28,6 +28,7 @@ #include "libavutil/internal.h" #include "libavutil/mem.h" #include "libavutil/intreadwrite.h" +#include "thread.h" #define READ_PIXELS(a, b, c) \ do { \ @@ -37,6 +38,12 @@ *c++ = (val >> 20) & 0x3FF; \ } while (0) +typedef struct ThreadData { + AVFrame *frame; + uint8_t *buf; + int stride; +} ThreadData; + static void v210_planar_unpack_c(const uint32_t *src, uint16_t *y, uint16_t *u, uint16_t *v, int width) { uint32_t val; @@ -64,21 +71,81 @@ static av_cold int decode_init(AVCodecContext *avctx) avctx->pix_fmt = AV_PIX_FMT_YUV422P10; avctx->bits_per_raw_sample = 10; + s->thread_count = av_clip(avctx->thread_count, 1, avctx->height/4); s->aligned_input = 0; ff_v210dec_init(s); return 0; } +static int v210_decode_slice(AVCodecContext *avctx, void *arg, int jobnr, int threadnr) +{ + V210DecContext *s = avctx->priv_data; + int h, w; + ThreadData *td = arg; + AVFrame *frame = td->frame; + int stride = td->stride; + int slice_start = (avctx->height * jobnr) / s->thread_count; + int slice_end = (avctx->height * (jobnr+1)) / s->thread_count; + uint8_t *psrc = td->buf + stride * slice_start; + uint16_t *y, *u, *v; + + y = (uint16_t*)frame->data[0] + slice_start * frame->linesize[0] / 2; + u = (uint16_t*)frame->data[1] + slice_start * frame->linesize[1] / 2; + v = (uint16_t*)frame->data[2] + slice_start * frame->linesize[2] / 2; + for (h = slice_start; h < slice_end; h++) { + const uint32_t *src = (const uint32_t*)psrc; + uint32_t val; + + w = (avctx->width / 12) * 12; + s->unpack_frame(src, y, u, v, w); + + y += w; + u += w >> 1; + v += w >> 1; + src += (w << 1) / 3; + + if (w < avctx->width - 5) { + READ_PIXELS(u, y, v); + READ_PIXELS(y, u, y); + READ_PIXELS(v, y, u); + READ_PIXELS(y, v, y); + w += 6; + } + + if (w < avctx->width - 1) { + READ_PIXELS(u, y, v); + + val = av_le2ne32(*src++); + *y++ = val & 0x3FF; + if (w < avctx->width - 3) { + *u++ = (val >> 10) & 0x3FF; + *y++ = (val >> 20) & 0x3FF; + + val = av_le2ne32(*src++); + *v++ = val & 0x3FF; + *y++ = (val >> 10) & 0x3FF; + } + } + + psrc += stride; + y += frame->linesize[0] / 2 - avctx->width + (avctx->width & 1); + u += frame->linesize[1] / 2 - avctx->width / 2; + v += frame->linesize[2] / 2 - avctx->width / 2; + } + + return 0; +} + static int decode_frame(AVCodecContext *avctx, void *data, int *got_frame, AVPacket *avpkt) { V210DecContext *s = avctx->priv_data; - - int h, w, ret, stride, aligned_input; + ThreadData td; + int ret, stride, aligned_input; + ThreadFrame frame = { .f = data }; AVFrame *pic = data; const uint8_t *psrc = avpkt->data; - uint16_t *y, *u, *v; if (s->custom_stride ) stride = s->custom_stride; @@ -86,6 +153,7 @@ static int decode_frame(AVCodecContext *avctx, void *data, int *got_frame, int aligned_width = ((avctx->width + 47) / 48) * 48; stride = aligned_width * 8 / 3; } + td.stride = stride; if (avpkt->size < stride * avctx->height) { if ((((avctx->width + 23) / 24) * 24 * 8) / 3 * avctx->height == avpkt->size) { @@ -110,55 +178,15 @@ static int decode_frame(AVCodecContext *avctx, void *data, int *got_frame, ff_v210dec_init(s); } - if ((ret = ff_get_buffer(avctx, pic, 0)) < 0) + if ((ret = ff_thread_get_buffer(avctx, &frame, 0)) < 0) return ret; - y = (uint16_t*)pic->data[0]; - u = (uint16_t*)pic->data[1]; - v = (uint16_t*)pic->data[2]; pic->pict_type = AV_PICTURE_TYPE_I; pic->key_frame = 1; - for (h = 0; h < avctx->height; h++) { - const uint32_t *src = (const uint32_t*)psrc; - uint32_t val; - - w = (avctx->width / 12) * 12; - s->unpack_frame(src, y, u, v, w); - - y += w; - u += w >> 1; - v += w >> 1; - src += (w << 1) / 3; - - if (w < avctx->width - 5) { - READ_PIXELS(u, y, v); - READ_PIXELS(y, u, y); - READ_PIXELS(v, y, u); - READ_PIXELS(y, v, y); - w += 6; - } - - if (w < avctx->width - 1) { - READ_PIXELS(u, y, v); - - val = av_le2ne32(*src++); - *y++ = val & 0x3FF; - if (w < avctx->width - 3) { - *u++ = (val >> 10) & 0x3FF; - *y++ = (val >> 20) & 0x3FF; - - val = av_le2ne32(*src++); - *v++ = val & 0x3FF; - *y++ = (val >> 10) & 0x3FF; - } - } - - psrc += stride; - y += pic->linesize[0] / 2 - avctx->width + (avctx->width & 1); - u += pic->linesize[1] / 2 - avctx->width / 2; - v += pic->linesize[2] / 2 - avctx->width / 2; - } + td.buf = (uint8_t*)psrc; + td.frame = pic; + avctx->execute2(avctx, v210_decode_slice, &td, NULL, s->thread_count); if (avctx->field_order > AV_FIELD_PROGRESSIVE) { /* we have interlaced material flagged in container */ @@ -194,6 +222,8 @@ AVCodec ff_v210_decoder = { .priv_data_size = sizeof(V210DecContext), .init = decode_init, .decode = decode_frame, - .capabilities = AV_CODEC_CAP_DR1, + .capabilities = AV_CODEC_CAP_DR1 | + AV_CODEC_CAP_SLICE_THREADS | + AV_CODEC_CAP_FRAME_THREADS, .priv_class = &v210dec_class, }; diff --git a/libavcodec/v210dec.h b/libavcodec/v210dec.h index cfdb29da09..662e266315 100644 --- a/libavcodec/v210dec.h +++ b/libavcodec/v210dec.h @@ -27,6 +27,7 @@ typedef struct { AVClass *av_class; int custom_stride; int aligned_input; + int thread_count; int stride_warning_shown; void (*unpack_frame)(const uint32_t *src, uint16_t *y, uint16_t *u, uint16_t *v, int width); } V210DecContext;

1 0

avcodec/v410dec: add support for frame and slice threading
by Limin Wang 25 Nov '19

25 Nov '19

ffmpeg | branch: master | Limin Wang <lance.lmwang(a)gmail.com> | Mon Nov 25 11:40:04 2019 +0800| [f0dbeb5eaa41fa508560ddaee51efa875a602bfc] | committer: Michael Niedermayer avcodec/v410dec: add support for frame and slice threading 1, Test server configure: [root@localhost ~]# cat /proc/cpuinfo |grep "model name" model name : Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz model name : Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz ... [root@localhost ~]# free -h total used free shared buff/cache available Mem: 102G 1.1G 100G 16M 657M 100G Swap: 4.0G 0B 4.0G 2, Test result: encode the v410 input data for testing: ./ffmpeg -y -i 4k_422.ts -c:v v410 -vframes 10 test.avi master: ./ffmpeg -y -stream_loop 1000 -i ./test.avi -benchmark -f null - frame=10010 fps= 37 q=-0.0 Lsize=N/A time=00:38:26.30 bitrate=N/A speed= 8.6x video:5240kB audio:432432kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown bench: utime=166.016s stime=102.192s rtime=268.120s bench: maxrss=273400kB patch applied: ./ffmpeg -y -threads 2 -thread_type slice -stream_loop 1000 -i ./test.avi -benchmark -f null - frame=10010 fps= 53 q=-0.0 Lsize=N/A time=00:38:26.30 bitrate=N/A speed=12.3x video:5240kB audio:432432kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown bench: utime=165.135s stime=100.456s rtime=187.994s bench: maxrss=275476kB ./ffmpeg -y -threads 2 -thread_type frame -stream_loop 1000 -i ./test.avi -benchmark -f null - frame=10010 fps= 61 q=-0.0 Lsize=N/A time=00:38:26.30 bitrate=N/A speed=14.1x video:5240kB audio:432432kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown bench: utime=171.386s stime=122.102s rtime=163.637s bench: maxrss=340308kB Signed-off-by: Limin Wang <lance.lmwang(a)gmail.com> Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=f0dbeb5eaa41fa508… --- libavcodec/v410dec.c | 72 +++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 51 insertions(+), 21 deletions(-) diff --git a/libavcodec/v410dec.c b/libavcodec/v410dec.c index 48fab68273..7ad5eb8fb5 100644 --- a/libavcodec/v410dec.c +++ b/libavcodec/v410dec.c @@ -24,6 +24,13 @@ #include "libavutil/intreadwrite.h" #include "avcodec.h" #include "internal.h" +#include "thread.h" + +typedef struct ThreadData { + AVFrame *frame; + uint8_t *buf; + int stride; +} ThreadData; static av_cold int v410_decode_init(AVCodecContext *avctx) { @@ -42,31 +49,24 @@ static av_cold int v410_decode_init(AVCodecContext *avctx) return 0; } -static int v410_decode_frame(AVCodecContext *avctx, void *data, - int *got_frame, AVPacket *avpkt) +static int v410_decode_slice(AVCodecContext *avctx, void *arg, int jobnr, int threadnr) { - AVFrame *pic = data; - uint8_t *src = avpkt->data; + ThreadData *td = arg; + AVFrame *pic = td->frame; + int stride = td->stride; + int thread_count = av_clip(avctx->thread_count, 1, avctx->height/4); + int slice_start = (avctx->height * jobnr) / thread_count; + int slice_end = (avctx->height * (jobnr+1)) / thread_count; + const uint8_t *src = td->buf + stride * slice_start; uint16_t *y, *u, *v; uint32_t val; - int i, j, ret; - - if (avpkt->size < 4 * avctx->height * avctx->width) { - av_log(avctx, AV_LOG_ERROR, "Insufficient input data.\n"); - return AVERROR(EINVAL); - } - - if ((ret = ff_get_buffer(avctx, pic, 0)) < 0) - return ret; + int i, j; - pic->key_frame = 1; - pic->pict_type = AV_PICTURE_TYPE_I; + y = (uint16_t*)pic->data[0] + slice_start * (pic->linesize[0] >> 1); + u = (uint16_t*)pic->data[1] + slice_start * (pic->linesize[1] >> 1); + v = (uint16_t*)pic->data[2] + slice_start * (pic->linesize[2] >> 1); - y = (uint16_t *)pic->data[0]; - u = (uint16_t *)pic->data[1]; - v = (uint16_t *)pic->data[2]; - - for (i = 0; i < avctx->height; i++) { + for (i = slice_start; i < slice_end; i++) { for (j = 0; j < avctx->width; j++) { val = AV_RL32(src); @@ -82,6 +82,35 @@ static int v410_decode_frame(AVCodecContext *avctx, void *data, v += pic->linesize[2] >> 1; } + return 0; +} + +static int v410_decode_frame(AVCodecContext *avctx, void *data, + int *got_frame, AVPacket *avpkt) +{ + ThreadData td; + ThreadFrame frame = { .f = data }; + AVFrame *pic = data; + uint8_t *src = avpkt->data; + int ret; + int thread_count = av_clip(avctx->thread_count, 1, avctx->height/4); + + td.stride = avctx->width * 4; + if (avpkt->size < 4 * avctx->height * avctx->width) { + av_log(avctx, AV_LOG_ERROR, "Insufficient input data.\n"); + return AVERROR(EINVAL); + } + + if ((ret = ff_thread_get_buffer(avctx, &frame, 0)) < 0) + return ret; + + pic->key_frame = 1; + pic->pict_type = AV_PICTURE_TYPE_I; + + td.buf = src; + td.frame = pic; + avctx->execute2(avctx, v410_decode_slice, &td, NULL, thread_count); + *got_frame = 1; return avpkt->size; @@ -94,5 +123,6 @@ AVCodec ff_v410_decoder = { .id = AV_CODEC_ID_V410, .init = v410_decode_init, .decode = v410_decode_frame, - .capabilities = AV_CODEC_CAP_DR1, + .capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_SLICE_THREADS | + AV_CODEC_CAP_FRAME_THREADS };

1 0