FFmpeg
libaomdec.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2010, Google, Inc.
3  *
4  * This file is part of FFmpeg.
5  *
6  * FFmpeg is free software; you can redistribute it and/or
7  * modify it under the terms of the GNU Lesser General Public
8  * License as published by the Free Software Foundation; either
9  * version 2.1 of the License, or (at your option) any later version.
10  *
11  * FFmpeg is distributed in the hope that it will be useful,
12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14  * Lesser General Public License for more details.
15  *
16  * You should have received a copy of the GNU Lesser General Public
17  * License along with FFmpeg; if not, write to the Free Software
18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19  */
20 
21 /**
22  * @file
23  * AV1 decoder support via libaom
24  */
25 
26 #include <aom/aom_decoder.h>
27 #include <aom/aomdx.h>
28 
29 #include "libavutil/common.h"
30 #include "libavutil/imgutils.h"
31 
32 #include "avcodec.h"
33 #include "internal.h"
34 #include "profiles.h"
35 
36 typedef struct AV1DecodeContext {
37  struct aom_codec_ctx decoder;
39 
40 static av_cold int aom_init(AVCodecContext *avctx,
41  const struct aom_codec_iface *iface)
42 {
43  AV1DecodeContext *ctx = avctx->priv_data;
44  struct aom_codec_dec_cfg deccfg = {
45  .threads = FFMIN(avctx->thread_count ? avctx->thread_count : av_cpu_count(), 16)
46  };
47 
48  av_log(avctx, AV_LOG_INFO, "%s\n", aom_codec_version_str());
49  av_log(avctx, AV_LOG_VERBOSE, "%s\n", aom_codec_build_config());
50 
51  if (aom_codec_dec_init(&ctx->decoder, iface, &deccfg, 0) != AOM_CODEC_OK) {
52  const char *error = aom_codec_error(&ctx->decoder);
53  av_log(avctx, AV_LOG_ERROR, "Failed to initialize decoder: %s\n",
54  error);
55  return AVERROR(EINVAL);
56  }
57 
58  return 0;
59 }
60 
61 static void image_copy_16_to_8(AVFrame *pic, struct aom_image *img)
62 {
64  int i;
65 
66  for (i = 0; i < desc->nb_components; i++) {
67  int w = img->d_w;
68  int h = img->d_h;
69  int x, y;
70 
71  if (i) {
72  w = (w + img->x_chroma_shift) >> img->x_chroma_shift;
73  h = (h + img->y_chroma_shift) >> img->y_chroma_shift;
74  }
75 
76  for (y = 0; y < h; y++) {
77  uint16_t *src = (uint16_t *)(img->planes[i] + y * img->stride[i]);
78  uint8_t *dst = pic->data[i] + y * pic->linesize[i];
79  for (x = 0; x < w; x++)
80  *dst++ = *src++;
81  }
82  }
83 }
84 
85 // returns 0 on success, AVERROR_INVALIDDATA otherwise
86 static int set_pix_fmt(AVCodecContext *avctx, struct aom_image *img)
87 {
88  static const enum AVColorRange color_ranges[] = {
90  };
91  avctx->color_range = color_ranges[img->range];
92  avctx->color_primaries = img->cp;
93  avctx->colorspace = img->mc;
94  avctx->color_trc = img->tc;
95 
96  switch (img->fmt) {
97  case AOM_IMG_FMT_I420:
98  case AOM_IMG_FMT_I42016:
99  if (img->bit_depth == 8) {
100  avctx->pix_fmt = img->monochrome ?
102  avctx->profile = FF_PROFILE_AV1_MAIN;
103  return 0;
104  } else if (img->bit_depth == 10) {
105  avctx->pix_fmt = img->monochrome ?
107  avctx->profile = FF_PROFILE_AV1_MAIN;
108  return 0;
109  } else if (img->bit_depth == 12) {
110  avctx->pix_fmt = img->monochrome ?
113  return 0;
114  } else {
115  return AVERROR_INVALIDDATA;
116  }
117  case AOM_IMG_FMT_I422:
118  case AOM_IMG_FMT_I42216:
119  if (img->bit_depth == 8) {
120  avctx->pix_fmt = AV_PIX_FMT_YUV422P;
122  return 0;
123  } else if (img->bit_depth == 10) {
124  avctx->pix_fmt = AV_PIX_FMT_YUV422P10;
126  return 0;
127  } else if (img->bit_depth == 12) {
128  avctx->pix_fmt = AV_PIX_FMT_YUV422P12;
130  return 0;
131  } else {
132  return AVERROR_INVALIDDATA;
133  }
134  case AOM_IMG_FMT_I444:
135  case AOM_IMG_FMT_I44416:
136  if (img->bit_depth == 8) {
137  avctx->pix_fmt = AV_PIX_FMT_YUV444P;
138  avctx->profile = FF_PROFILE_AV1_HIGH;
139  return 0;
140  } else if (img->bit_depth == 10) {
141  avctx->pix_fmt = AV_PIX_FMT_YUV444P10;
142  avctx->profile = FF_PROFILE_AV1_HIGH;
143  return 0;
144  } else if (img->bit_depth == 12) {
145  avctx->pix_fmt = AV_PIX_FMT_YUV444P12;
147  return 0;
148  } else {
149  return AVERROR_INVALIDDATA;
150  }
151 
152  default:
153  return AVERROR_INVALIDDATA;
154  }
155 }
156 
157 static int aom_decode(AVCodecContext *avctx, void *data, int *got_frame,
158  AVPacket *avpkt)
159 {
160  AV1DecodeContext *ctx = avctx->priv_data;
161  AVFrame *picture = data;
162  const void *iter = NULL;
163  struct aom_image *img;
164  int ret;
165 
166  if (aom_codec_decode(&ctx->decoder, avpkt->data, avpkt->size, NULL) !=
167  AOM_CODEC_OK) {
168  const char *error = aom_codec_error(&ctx->decoder);
169  const char *detail = aom_codec_error_detail(&ctx->decoder);
170 
171  av_log(avctx, AV_LOG_ERROR, "Failed to decode frame: %s\n", error);
172  if (detail)
173  av_log(avctx, AV_LOG_ERROR, " Additional information: %s\n",
174  detail);
175  return AVERROR_INVALIDDATA;
176  }
177 
178  if ((img = aom_codec_get_frame(&ctx->decoder, &iter))) {
179  if (img->d_w > img->w || img->d_h > img->h) {
180  av_log(avctx, AV_LOG_ERROR, "Display dimensions %dx%d exceed storage %dx%d\n",
181  img->d_w, img->d_h, img->w, img->h);
182  return AVERROR_EXTERNAL;
183  }
184 
185  if ((ret = set_pix_fmt(avctx, img)) < 0) {
186  av_log(avctx, AV_LOG_ERROR, "Unsupported output colorspace (%d) / bit_depth (%d)\n",
187  img->fmt, img->bit_depth);
188  return ret;
189  }
190 
191  if ((int)img->d_w != avctx->width || (int)img->d_h != avctx->height) {
192  av_log(avctx, AV_LOG_INFO, "dimension change! %dx%d -> %dx%d\n",
193  avctx->width, avctx->height, img->d_w, img->d_h);
194  ret = ff_set_dimensions(avctx, img->d_w, img->d_h);
195  if (ret < 0)
196  return ret;
197  }
198  if ((ret = ff_get_buffer(avctx, picture, 0)) < 0)
199  return ret;
200 
201 #ifdef AOM_CTRL_AOMD_GET_FRAME_FLAGS
202  {
203  aom_codec_frame_flags_t flags;
204  ret = aom_codec_control(&ctx->decoder, AOMD_GET_FRAME_FLAGS, &flags);
205  if (ret == AOM_CODEC_OK) {
206  picture->key_frame = !!(flags & AOM_FRAME_IS_KEY);
207  if (flags & (AOM_FRAME_IS_KEY | AOM_FRAME_IS_INTRAONLY))
208  picture->pict_type = AV_PICTURE_TYPE_I;
209  else if (flags & AOM_FRAME_IS_SWITCH)
210  picture->pict_type = AV_PICTURE_TYPE_SP;
211  else
212  picture->pict_type = AV_PICTURE_TYPE_P;
213  }
214  }
215 #endif
216 
218  &picture->sample_aspect_ratio.den,
219  picture->height * img->r_w,
220  picture->width * img->r_h,
221  INT_MAX);
222  ff_set_sar(avctx, picture->sample_aspect_ratio);
223 
224  if ((img->fmt & AOM_IMG_FMT_HIGHBITDEPTH) && img->bit_depth == 8)
225  image_copy_16_to_8(picture, img);
226  else
227  av_image_copy(picture->data, picture->linesize, (const uint8_t **)img->planes,
228  img->stride, avctx->pix_fmt, img->d_w, img->d_h);
229  *got_frame = 1;
230  }
231  return avpkt->size;
232 }
233 
234 static av_cold int aom_free(AVCodecContext *avctx)
235 {
236  AV1DecodeContext *ctx = avctx->priv_data;
237  aom_codec_destroy(&ctx->decoder);
238  return 0;
239 }
240 
241 static av_cold int av1_init(AVCodecContext *avctx)
242 {
243  return aom_init(avctx, &aom_codec_av1_dx_algo);
244 }
245 
247  .name = "libaom-av1",
248  .long_name = NULL_IF_CONFIG_SMALL("libaom AV1"),
249  .type = AVMEDIA_TYPE_VIDEO,
250  .id = AV_CODEC_ID_AV1,
251  .priv_data_size = sizeof(AV1DecodeContext),
252  .init = av1_init,
253  .close = aom_free,
254  .decode = aom_decode,
256  .caps_internal = FF_CODEC_CAP_AUTO_THREADS,
258  .wrapper_name = "libaom",
259 };
error
static void error(const char *err)
Definition: target_bsf_fuzzer.c:30
AVCodec
AVCodec.
Definition: codec.h:197
aom_free
static av_cold int aom_free(AVCodecContext *avctx)
Definition: libaomdec.c:234
aom_init
static av_cold int aom_init(AVCodecContext *avctx, const struct aom_codec_iface *iface)
Definition: libaomdec.c:40
init
static av_cold int init(AVCodecContext *avctx)
Definition: avrndec.c:31
AVERROR
Filter the word “frame” indicates either a video frame or a group of audio as stored in an AVFrame structure Format for each input and each output the list of supported formats For video that means pixel format For audio that means channel sample they are references to shared objects When the negotiation mechanism computes the intersection of the formats supported at each end of a all references to both lists are replaced with a reference to the intersection And when a single format is eventually chosen for a link amongst the remaining all references to the list are updated That means that if a filter requires that its input and output have the same format amongst a supported all it has to do is use a reference to the same list of formats query_formats can leave some formats unset and return AVERROR(EAGAIN) to cause the negotiation mechanism toagain later. That can be used by filters with complex requirements to use the format negotiated on one link to set the formats supported on another. Frame references ownership and permissions
AVCodecContext::colorspace
enum AVColorSpace colorspace
YUV colorspace type.
Definition: avcodec.h:1066
av_pix_fmt_desc_get
const AVPixFmtDescriptor * av_pix_fmt_desc_get(enum AVPixelFormat pix_fmt)
Definition: pixdesc.c:2541
AVFrame
This structure describes decoded (raw) audio or video data.
Definition: frame.h:303
AVCodecContext::color_trc
enum AVColorTransferCharacteristic color_trc
Color Transfer Characteristic.
Definition: avcodec.h:1059
AVFrame::width
int width
Definition: frame.h:361
w
uint8_t w
Definition: llviddspenc.c:39
AVCOL_RANGE_JPEG
@ AVCOL_RANGE_JPEG
Full range content.
Definition: pixfmt.h:576
internal.h
AVPacket::data
uint8_t * data
Definition: packet.h:365
set_pix_fmt
static int set_pix_fmt(AVCodecContext *avctx, struct aom_image *img)
Definition: libaomdec.c:86
data
const char data[16]
Definition: mxf.c:142
AV_PIX_FMT_YUV420P10
#define AV_PIX_FMT_YUV420P10
Definition: pixfmt.h:389
ff_libaom_av1_decoder
const AVCodec ff_libaom_av1_decoder
Definition: libaomdec.c:246
AV_LOG_VERBOSE
#define AV_LOG_VERBOSE
Detailed information.
Definition: log.h:197
AVFrame::data
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:317
AVCodecContext::thread_count
int thread_count
thread count is used to decide how many independent tasks should be passed to execute()
Definition: avcodec.h:1546
AVFrame::key_frame
int key_frame
1 -> keyframe, 0-> not
Definition: frame.h:381
av_reduce
int av_reduce(int *dst_num, int *dst_den, int64_t num, int64_t den, int64_t max)
Reduce a fraction.
Definition: rational.c:35
AVRational::num
int num
Numerator.
Definition: rational.h:59
AV_PIX_FMT_YUV444P10
#define AV_PIX_FMT_YUV444P10
Definition: pixfmt.h:392
AV1DecodeContext::decoder
struct aom_codec_ctx decoder
Definition: libaomdec.c:37
AVCodecContext::color_primaries
enum AVColorPrimaries color_primaries
Chromaticity coordinates of the source primaries.
Definition: avcodec.h:1052
AV_LOG_ERROR
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
Definition: log.h:181
av_cold
#define av_cold
Definition: attributes.h:90
decode
static void decode(AVCodecContext *dec_ctx, AVPacket *pkt, AVFrame *frame, FILE *outfile)
Definition: decode_audio.c:71
AV_CODEC_CAP_OTHER_THREADS
#define AV_CODEC_CAP_OTHER_THREADS
Codec supports multithreading through a method other than slice- or frame-level multithreading.
Definition: codec.h:122
ctx
AVFormatContext * ctx
Definition: movenc.c:48
ff_av1_profiles
const AVProfile ff_av1_profiles[]
Definition: profiles.c:147
AV_PIX_FMT_YUV420P
@ AV_PIX_FMT_YUV420P
planar YUV 4:2:0, 12bpp, (1 Cr & Cb sample per 2x2 Y samples)
Definition: pixfmt.h:66
AV_PIX_FMT_GRAY10
#define AV_PIX_FMT_GRAY10
Definition: pixfmt.h:370
if
if(ret)
Definition: filter_design.txt:179
NULL
#define NULL
Definition: coverity.c:32
AVCodecContext::color_range
enum AVColorRange color_range
MPEG vs JPEG YUV range.
Definition: avcodec.h:1073
AV_CODEC_ID_AV1
@ AV_CODEC_ID_AV1
Definition: codec_id.h:278
AV_PICTURE_TYPE_I
@ AV_PICTURE_TYPE_I
Intra.
Definition: avutil.h:274
profiles.h
src
#define src
Definition: vp8dsp.c:255
AV_PIX_FMT_YUV422P10
#define AV_PIX_FMT_YUV422P10
Definition: pixfmt.h:390
AV_PIX_FMT_GRAY8
@ AV_PIX_FMT_GRAY8
Y , 8bpp.
Definition: pixfmt.h:74
AV_PICTURE_TYPE_SP
@ AV_PICTURE_TYPE_SP
Switching Predicted.
Definition: avutil.h:279
aom_decode
static int aom_decode(AVCodecContext *avctx, void *data, int *got_frame, AVPacket *avpkt)
Definition: libaomdec.c:157
av_cpu_count
int av_cpu_count(void)
Definition: cpu.c:184
FF_CODEC_CAP_AUTO_THREADS
#define FF_CODEC_CAP_AUTO_THREADS
Codec handles avctx->thread_count == 0 (auto) internally.
Definition: internal.h:80
AVFrame::pict_type
enum AVPictureType pict_type
Picture type of the frame.
Definition: frame.h:386
ff_get_buffer
int ff_get_buffer(AVCodecContext *avctx, AVFrame *frame, int flags)
Get a buffer for a frame.
Definition: decode.c:1638
FF_PROFILE_AV1_MAIN
#define FF_PROFILE_AV1_MAIN
Definition: avcodec.h:1727
AV_CODEC_CAP_DR1
#define AV_CODEC_CAP_DR1
Codec uses get_buffer() or get_encode_buffer() for allocating buffers and supports custom allocators.
Definition: codec.h:52
AVPacket::size
int size
Definition: packet.h:366
NULL_IF_CONFIG_SMALL
#define NULL_IF_CONFIG_SMALL(x)
Return NULL if CONFIG_SMALL is true, otherwise the argument without modification.
Definition: internal.h:117
av1_init
static av_cold int av1_init(AVCodecContext *avctx)
Definition: libaomdec.c:241
AV_PIX_FMT_YUV422P12
#define AV_PIX_FMT_YUV422P12
Definition: pixfmt.h:394
AV_PIX_FMT_YUV444P12
#define AV_PIX_FMT_YUV444P12
Definition: pixfmt.h:396
AVFrame::format
int format
format of the frame, -1 if unknown or unset Values correspond to enum AVPixelFormat for video frames,...
Definition: frame.h:376
FFMIN
#define FFMIN(a, b)
Definition: common.h:105
img
#define img
Definition: vf_colormatrix.c:116
AVERROR_EXTERNAL
#define AVERROR_EXTERNAL
Generic error in an external library.
Definition: error.h:57
AV_LOG_INFO
#define AV_LOG_INFO
Standard information.
Definition: log.h:192
i
int i
Definition: input.c:407
common.h
AVCodec::name
const char * name
Name of the codec implementation.
Definition: codec.h:204
AVCodecContext::height
int height
Definition: avcodec.h:674
AVCodecContext::pix_fmt
enum AVPixelFormat pix_fmt
Pixel format, see AV_PIX_FMT_xxx.
Definition: avcodec.h:711
AVCOL_RANGE_MPEG
@ AVCOL_RANGE_MPEG
Narrow or limited range content.
Definition: pixfmt.h:559
FF_PROFILE_AV1_PROFESSIONAL
#define FF_PROFILE_AV1_PROFESSIONAL
Definition: avcodec.h:1729
avcodec.h
ret
ret
Definition: filter_design.txt:187
AVFrame::sample_aspect_ratio
AVRational sample_aspect_ratio
Sample aspect ratio for the video frame, 0/1 if unknown/unspecified.
Definition: frame.h:391
AV_PIX_FMT_YUV420P12
#define AV_PIX_FMT_YUV420P12
Definition: pixfmt.h:393
image_copy_16_to_8
static void image_copy_16_to_8(AVFrame *pic, struct aom_image *img)
Definition: libaomdec.c:61
ff_set_sar
int ff_set_sar(AVCodecContext *avctx, AVRational sar)
Check that the provided sample aspect ratio is valid and set it on the codec context.
Definition: utils.c:99
AVCodecContext
main external API structure.
Definition: avcodec.h:501
AVFrame::height
int height
Definition: frame.h:361
av_image_copy
void av_image_copy(uint8_t *dst_data[4], int dst_linesizes[4], const uint8_t *src_data[4], const int src_linesizes[4], enum AVPixelFormat pix_fmt, int width, int height)
Copy image in src_data to dst_data.
Definition: imgutils.c:422
AVRational::den
int den
Denominator.
Definition: rational.h:60
AVCodecContext::profile
int profile
profile
Definition: avcodec.h:1631
AV1DecodeContext
Definition: libaomdec.c:36
profiles
static const AVProfile profiles[]
Definition: libfdk-aacenc.c:427
AV_PIX_FMT_YUV444P
@ AV_PIX_FMT_YUV444P
planar YUV 4:4:4, 24bpp, (1 Cr & Cb sample per 1x1 Y samples)
Definition: pixfmt.h:71
desc
const char * desc
Definition: libsvtav1.c:79
AV_PICTURE_TYPE_P
@ AV_PICTURE_TYPE_P
Predicted.
Definition: avutil.h:275
AVMEDIA_TYPE_VIDEO
@ AVMEDIA_TYPE_VIDEO
Definition: avutil.h:201
AV_PIX_FMT_YUV422P
@ AV_PIX_FMT_YUV422P
planar YUV 4:2:2, 16bpp, (1 Cr & Cb sample per 2x1 Y samples)
Definition: pixfmt.h:70
ff_set_dimensions
int ff_set_dimensions(AVCodecContext *s, int width, int height)
Check that the provided frame dimensions are valid and set them on the codec context.
Definition: utils.c:84
AVPixFmtDescriptor
Descriptor that unambiguously describes how the bits of a pixel are stored in the up to 4 data planes...
Definition: pixdesc.h:70
AVPacket
This structure stores compressed data.
Definition: packet.h:342
AVCodecContext::priv_data
void * priv_data
Definition: avcodec.h:528
AVCodecContext::width
int width
picture width / height.
Definition: avcodec.h:674
imgutils.h
flags
#define flags(name, subs,...)
Definition: cbs_av1.c:561
AVFrame::linesize
int linesize[AV_NUM_DATA_POINTERS]
For video, size in bytes of each picture line.
Definition: frame.h:334
av_log
#define av_log(a,...)
Definition: tableprint_vlc.h:28
AVERROR_INVALIDDATA
#define AVERROR_INVALIDDATA
Invalid data found when processing input.
Definition: error.h:59
FF_PROFILE_AV1_HIGH
#define FF_PROFILE_AV1_HIGH
Definition: avcodec.h:1728
h
h
Definition: vp9dsp_template.c:2038
AV_PIX_FMT_GRAY12
#define AV_PIX_FMT_GRAY12
Definition: pixfmt.h:371
AVColorRange
AVColorRange
Visual content value range.
Definition: pixfmt.h:541