FFmpeg
rtpdec_hevc.c
Go to the documentation of this file.
1 /*
2  * RTP parser for HEVC/H.265 payload format (draft version 6)
3  * Copyright (c) 2014 Thomas Volkert <thomas@homer-conferencing.com>
4  *
5  * This file is part of FFmpeg.
6  *
7  * FFmpeg is free software; you can redistribute it and/or
8  * modify it under the terms of the GNU Lesser General Public
9  * License as published by the Free Software Foundation; either
10  * version 2.1 of the License, or (at your option) any later version.
11  *
12  * FFmpeg is distributed in the hope that it will be useful,
13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15  * Lesser General Public License for more details.
16  *
17  * You should have received a copy of the GNU Lesser General Public
18  * License along with FFmpeg; if not, write to the Free Software
19  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20  */
21 
22 #include "libavutil/avassert.h"
23 #include "libavutil/avstring.h"
24 #include "libavutil/base64.h"
25 #include "libavcodec/get_bits.h"
26 
27 #include "avformat.h"
28 #include "internal.h"
29 #include "rtpdec.h"
30 #include "rtpdec_formats.h"
31 
32 #define RTP_HEVC_PAYLOAD_HEADER_SIZE 2
33 #define RTP_HEVC_FU_HEADER_SIZE 1
34 #define RTP_HEVC_DONL_FIELD_SIZE 2
35 #define RTP_HEVC_DOND_FIELD_SIZE 1
36 #define RTP_HEVC_AP_NALU_LENGTH_FIELD_SIZE 2
37 #define HEVC_SPECIFIED_NAL_UNIT_TYPES 48
38 
39 /* SDP out-of-band signaling data */
40 struct PayloadContext {
43  uint8_t *sps, *pps, *vps, *sei;
45 };
46 
47 static const uint8_t start_sequence[] = { 0x00, 0x00, 0x00, 0x01 };
48 
50  AVStream *stream,
51  PayloadContext *hevc_data,
52  const char *attr, const char *value)
53 {
54  /* profile-space: 0-3 */
55  /* profile-id: 0-31 */
56  if (!strcmp(attr, "profile-id")) {
57  hevc_data->profile_id = atoi(value);
58  av_log(s, AV_LOG_TRACE, "SDP: found profile-id: %d\n", hevc_data->profile_id);
59  }
60 
61  /* tier-flag: 0-1 */
62  /* level-id: 0-255 */
63  /* interop-constraints: [base16] */
64  /* profile-compatibility-indicator: [base16] */
65  /* sprop-sub-layer-id: 0-6, defines highest possible value for TID, default: 6 */
66  /* recv-sub-layer-id: 0-6 */
67  /* max-recv-level-id: 0-255 */
68  /* tx-mode: MSM,SSM */
69  /* sprop-vps: [base64] */
70  /* sprop-sps: [base64] */
71  /* sprop-pps: [base64] */
72  /* sprop-sei: [base64] */
73  if (!strcmp(attr, "sprop-vps") || !strcmp(attr, "sprop-sps") ||
74  !strcmp(attr, "sprop-pps") || !strcmp(attr, "sprop-sei")) {
75  uint8_t **data_ptr = NULL;
76  int *size_ptr = NULL;
77  if (!strcmp(attr, "sprop-vps")) {
78  data_ptr = &hevc_data->vps;
79  size_ptr = &hevc_data->vps_size;
80  } else if (!strcmp(attr, "sprop-sps")) {
81  data_ptr = &hevc_data->sps;
82  size_ptr = &hevc_data->sps_size;
83  } else if (!strcmp(attr, "sprop-pps")) {
84  data_ptr = &hevc_data->pps;
85  size_ptr = &hevc_data->pps_size;
86  } else if (!strcmp(attr, "sprop-sei")) {
87  data_ptr = &hevc_data->sei;
88  size_ptr = &hevc_data->sei_size;
89  } else
90  av_assert0(0);
91 
93  size_ptr, value);
94  }
95 
96  /* max-lsr, max-lps, max-cpb, max-dpb, max-br, max-tr, max-tc */
97  /* max-fps */
98 
99  /* sprop-max-don-diff: 0-32767
100 
101  When the RTP stream depends on one or more other RTP
102  streams (in this case tx-mode MUST be equal to "MSM" and
103  MSM is in use), this parameter MUST be present and the
104  value MUST be greater than 0.
105  */
106  if (!strcmp(attr, "sprop-max-don-diff")) {
107  if (atoi(value) > 0)
108  hevc_data->using_donl_field = 1;
109  av_log(s, AV_LOG_TRACE, "Found sprop-max-don-diff in SDP, DON field usage is: %d\n",
110  hevc_data->using_donl_field);
111  }
112 
113  /* sprop-depack-buf-nalus: 0-32767 */
114  if (!strcmp(attr, "sprop-depack-buf-nalus")) {
115  if (atoi(value) > 0)
116  hevc_data->using_donl_field = 1;
117  av_log(s, AV_LOG_TRACE, "Found sprop-depack-buf-nalus in SDP, DON field usage is: %d\n",
118  hevc_data->using_donl_field);
119  }
120 
121  /* sprop-depack-buf-bytes: 0-4294967295 */
122  /* depack-buf-cap */
123  /* sprop-segmentation-id: 0-3 */
124  /* sprop-spatial-segmentation-idc: [base16] */
125  /* dec-parallel-ca: */
126  /* include-dph */
127 
128  return 0;
129 }
130 
131 static av_cold int hevc_parse_sdp_line(AVFormatContext *ctx, int st_index,
132  PayloadContext *hevc_data, const char *line)
133 {
134  AVStream *current_stream;
135  AVCodecParameters *par;
136  const char *sdp_line_ptr = line;
137 
138  if (st_index < 0)
139  return 0;
140 
141  current_stream = ctx->streams[st_index];
142  par = current_stream->codecpar;
143 
144  if (av_strstart(sdp_line_ptr, "framesize:", &sdp_line_ptr)) {
145  ff_h264_parse_framesize(par, sdp_line_ptr);
146  } else if (av_strstart(sdp_line_ptr, "fmtp:", &sdp_line_ptr)) {
147  int ret = ff_parse_fmtp(ctx, current_stream, hevc_data, sdp_line_ptr,
149  if (hevc_data->vps_size || hevc_data->sps_size ||
150  hevc_data->pps_size || hevc_data->sei_size) {
151  par->extradata_size = hevc_data->vps_size + hevc_data->sps_size +
152  hevc_data->pps_size + hevc_data->sei_size;
153  if ((ret = ff_alloc_extradata(par, par->extradata_size)) >= 0) {
154  int pos = 0;
155  memcpy(par->extradata + pos, hevc_data->vps, hevc_data->vps_size);
156  pos += hevc_data->vps_size;
157  memcpy(par->extradata + pos, hevc_data->sps, hevc_data->sps_size);
158  pos += hevc_data->sps_size;
159  memcpy(par->extradata + pos, hevc_data->pps, hevc_data->pps_size);
160  pos += hevc_data->pps_size;
161  memcpy(par->extradata + pos, hevc_data->sei, hevc_data->sei_size);
162  }
163 
164  av_freep(&hevc_data->vps);
165  av_freep(&hevc_data->sps);
166  av_freep(&hevc_data->pps);
167  av_freep(&hevc_data->sei);
168  hevc_data->vps_size = 0;
169  hevc_data->sps_size = 0;
170  hevc_data->pps_size = 0;
171  hevc_data->sei_size = 0;
172  }
173  return ret;
174  }
175 
176  return 0;
177 }
178 
180  AVStream *st, AVPacket *pkt, uint32_t *timestamp,
181  const uint8_t *buf, int len, uint16_t seq,
182  int flags)
183 {
184  const uint8_t *rtp_pl = buf;
185  int tid, lid, nal_type;
186  int first_fragment, last_fragment, fu_type;
187  uint8_t new_nal_header[2];
188  int res = 0;
189 
190  /* sanity check for size of input packet: 1 byte payload at least */
191  if (len < RTP_HEVC_PAYLOAD_HEADER_SIZE + 1) {
192  av_log(ctx, AV_LOG_ERROR, "Too short RTP/HEVC packet, got %d bytes\n", len);
193  return AVERROR_INVALIDDATA;
194  }
195 
196  /*
197  * decode the HEVC payload header according to section 4 of draft version 6:
198  *
199  * 0 1
200  * 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
201  * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
202  * |F| Type | LayerId | TID |
203  * +-------------+-----------------+
204  *
205  * Forbidden zero (F): 1 bit
206  * NAL unit type (Type): 6 bits
207  * NUH layer ID (LayerId): 6 bits
208  * NUH temporal ID plus 1 (TID): 3 bits
209  */
210  nal_type = (buf[0] >> 1) & 0x3f;
211  lid = ((buf[0] << 5) & 0x20) | ((buf[1] >> 3) & 0x1f);
212  tid = buf[1] & 0x07;
213 
214  /* sanity check for correct layer ID */
215  if (lid) {
216  /* future scalable or 3D video coding extensions */
217  avpriv_report_missing_feature(ctx, "Multi-layer HEVC coding");
218  return AVERROR_PATCHWELCOME;
219  }
220 
221  /* sanity check for correct temporal ID */
222  if (!tid) {
223  av_log(ctx, AV_LOG_ERROR, "Illegal temporal ID in RTP/HEVC packet\n");
224  return AVERROR_INVALIDDATA;
225  }
226 
227  /* sanity check for correct NAL unit type */
228  if (nal_type > 50) {
229  av_log(ctx, AV_LOG_ERROR, "Unsupported (HEVC) NAL type (%d)\n", nal_type);
230  return AVERROR_INVALIDDATA;
231  }
232 
233  switch (nal_type) {
234  /* video parameter set (VPS) */
235  case 32:
236  /* sequence parameter set (SPS) */
237  case 33:
238  /* picture parameter set (PPS) */
239  case 34:
240  /* supplemental enhancement information (SEI) */
241  case 39:
242  /* single NAL unit packet */
243  default:
244  /* create A/V packet */
245  if ((res = av_new_packet(pkt, sizeof(start_sequence) + len)) < 0)
246  return res;
247  /* A/V packet: copy start sequence */
248  memcpy(pkt->data, start_sequence, sizeof(start_sequence));
249  /* A/V packet: copy NAL unit data */
250  memcpy(pkt->data + sizeof(start_sequence), buf, len);
251 
252  break;
253  /* aggregated packet (AP) - with two or more NAL units */
254  case 48:
255  /* pass the HEVC payload header */
258 
259  /* pass the HEVC DONL field */
260  if (rtp_hevc_ctx->using_donl_field) {
263  }
264 
265  res = ff_h264_handle_aggregated_packet(ctx, rtp_hevc_ctx, pkt, buf, len,
266  rtp_hevc_ctx->using_donl_field ?
268  NULL, 0);
269  if (res < 0)
270  return res;
271  break;
272  /* fragmentation unit (FU) */
273  case 49:
274  /* pass the HEVC payload header */
277 
278  /*
279  * decode the FU header
280  *
281  * 0 1 2 3 4 5 6 7
282  * +-+-+-+-+-+-+-+-+
283  * |S|E| FuType |
284  * +---------------+
285  *
286  * Start fragment (S): 1 bit
287  * End fragment (E): 1 bit
288  * FuType: 6 bits
289  */
290  first_fragment = buf[0] & 0x80;
291  last_fragment = buf[0] & 0x40;
292  fu_type = buf[0] & 0x3f;
293 
294  /* pass the HEVC FU header */
297 
298  /* pass the HEVC DONL field */
299  if (rtp_hevc_ctx->using_donl_field) {
302  }
303 
304  av_log(ctx, AV_LOG_TRACE, " FU type %d with %d bytes\n", fu_type, len);
305 
306  /* sanity check for size of input packet: 1 byte payload at least */
307  if (len <= 0) {
308  if (len < 0) {
310  "Too short RTP/HEVC packet, got %d bytes of NAL unit type %d\n",
311  len, nal_type);
312  return AVERROR_INVALIDDATA;
313  } else {
314  return AVERROR(EAGAIN);
315  }
316  }
317 
318  if (first_fragment && last_fragment) {
319  av_log(ctx, AV_LOG_ERROR, "Illegal combination of S and E bit in RTP/HEVC packet\n");
320  return AVERROR_INVALIDDATA;
321  }
322 
323  new_nal_header[0] = (rtp_pl[0] & 0x81) | (fu_type << 1);
324  new_nal_header[1] = rtp_pl[1];
325 
326  res = ff_h264_handle_frag_packet(pkt, buf, len, first_fragment,
327  new_nal_header, sizeof(new_nal_header));
328 
329  break;
330  /* PACI packet */
331  case 50:
332  /* Temporal scalability control information (TSCI) */
333  avpriv_report_missing_feature(ctx, "PACI packets for RTP/HEVC");
334  res = AVERROR_PATCHWELCOME;
335  break;
336  }
337 
338  pkt->stream_index = st->index;
339 
340  return res;
341 }
342 
344  .enc_name = "H265",
345  .codec_type = AVMEDIA_TYPE_VIDEO,
346  .codec_id = AV_CODEC_ID_HEVC,
347  .need_parsing = AVSTREAM_PARSE_FULL,
348  .priv_data_size = sizeof(PayloadContext),
349  .parse_sdp_a_line = hevc_parse_sdp_line,
351 };
AVCodecParameters::extradata
uint8_t * extradata
Extra binary data needed for initializing the decoder, codec-dependent.
Definition: codec_par.h:74
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
AVCodecParameters
This struct describes the properties of an encoded stream.
Definition: codec_par.h:52
rtpdec_formats.h
ff_parse_fmtp
int ff_parse_fmtp(AVFormatContext *s, AVStream *stream, PayloadContext *data, const char *p, int(*parse_fmtp)(AVFormatContext *s, AVStream *stream, PayloadContext *data, const char *attr, const char *value))
Definition: rtpdec.c:887
RTP_HEVC_PAYLOAD_HEADER_SIZE
#define RTP_HEVC_PAYLOAD_HEADER_SIZE
Definition: rtpdec_hevc.c:32
AVFormatContext::streams
AVStream ** streams
A list of all streams in the file.
Definition: avformat.h:1403
AVPacket::data
uint8_t * data
Definition: packet.h:355
PayloadContext::sps_size
int sps_size
Definition: rtpdec_hevc.c:44
PayloadContext::pps
uint8_t * pps
Definition: rtpdec_hevc.c:43
hevc_handle_packet
static int hevc_handle_packet(AVFormatContext *ctx, PayloadContext *rtp_hevc_ctx, AVStream *st, AVPacket *pkt, uint32_t *timestamp, const uint8_t *buf, int len, uint16_t seq, int flags)
Definition: rtpdec_hevc.c:179
RTP_HEVC_FU_HEADER_SIZE
#define RTP_HEVC_FU_HEADER_SIZE
Definition: rtpdec_hevc.c:33
RTPDynamicProtocolHandler::enc_name
const char * enc_name
Definition: rtpdec.h:116
PayloadContext::profile_id
int profile_id
Definition: rtpdec_hevc.c:42
start_sequence
static const uint8_t start_sequence[]
Definition: rtpdec_hevc.c:47
avassert.h
AV_LOG_TRACE
#define AV_LOG_TRACE
Extremely verbose debugging, useful for libav* development.
Definition: log.h:202
AV_LOG_ERROR
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
Definition: log.h:176
av_cold
#define av_cold
Definition: attributes.h:90
hevc_parse_sdp_line
static av_cold int hevc_parse_sdp_line(AVFormatContext *ctx, int st_index, PayloadContext *hevc_data, const char *line)
Definition: rtpdec_hevc.c:131
s
#define s(width, name)
Definition: cbs_vp9.c:257
av_new_packet
int av_new_packet(AVPacket *pkt, int size)
Allocate the payload of a packet and initialize its fields with default values.
Definition: avpacket.c:88
PayloadContext::sps
uint8_t * sps
Definition: rtpdec_hevc.c:43
RTP_HEVC_DONL_FIELD_SIZE
#define RTP_HEVC_DONL_FIELD_SIZE
Definition: rtpdec_hevc.c:34
av_assert0
#define av_assert0(cond)
assert() equivalent, that is always enabled.
Definition: avassert.h:37
ctx
AVFormatContext * ctx
Definition: movenc.c:48
get_bits.h
ff_h264_parse_framesize
void ff_h264_parse_framesize(AVCodecParameters *par, const char *p)
Definition: rtpdec_h264.c:184
parse_packet
static int parse_packet(AVFormatContext *s, AVPacket *pkt, int stream_index, int flush)
Parse a packet, add all split parts to parse_queue.
Definition: utils.c:1449
PayloadContext::vps_size
int vps_size
Definition: rtpdec_hevc.c:44
AVFormatContext
Format I/O context.
Definition: avformat.h:1335
internal.h
AVStream::codecpar
AVCodecParameters * codecpar
Codec parameters associated with this stream.
Definition: avformat.h:1012
NULL
#define NULL
Definition: coverity.c:32
AVERROR_PATCHWELCOME
#define AVERROR_PATCHWELCOME
Not yet implemented in FFmpeg, patches welcome.
Definition: error.h:62
base64.h
rtpdec.h
AVCodecParameters::extradata_size
int extradata_size
Size of the extradata content in bytes.
Definition: codec_par.h:78
PayloadContext::sei
uint8_t * sei
Definition: rtpdec_hevc.c:43
avpriv_report_missing_feature
void avpriv_report_missing_feature(void *avc, const char *msg,...) av_printf_format(2
Log a generic warning message about a missing feature.
line
Definition: graph2dot.c:48
av_strstart
int av_strstart(const char *str, const char *pfx, const char **ptr)
Return non-zero if pfx is a prefix of str.
Definition: avstring.c:34
ff_h264_handle_frag_packet
int ff_h264_handle_frag_packet(AVPacket *pkt, const uint8_t *buf, int len, int start_bit, const uint8_t *nal_header, int nal_header_len)
Definition: rtpdec_h264.c:264
ff_hevc_dynamic_handler
const RTPDynamicProtocolHandler ff_hevc_dynamic_handler
Definition: rtpdec_hevc.c:343
AV_CODEC_ID_HEVC
@ AV_CODEC_ID_HEVC
Definition: codec_id.h:223
value
it s the only field you need to keep assuming you have a context There is some magic you don t need to care about around this just let it vf default value
Definition: writing_filters.txt:86
uint8_t
uint8_t
Definition: audio_convert.c:194
PayloadContext::using_donl_field
int using_donl_field
Definition: rtpdec_hevc.c:41
len
int len
Definition: vorbis_enc_data.h:452
RTP_HEVC_DOND_FIELD_SIZE
#define RTP_HEVC_DOND_FIELD_SIZE
Definition: rtpdec_hevc.c:35
hevc_sdp_parse_fmtp_config
static av_cold int hevc_sdp_parse_fmtp_config(AVFormatContext *s, AVStream *stream, PayloadContext *hevc_data, const char *attr, const char *value)
Definition: rtpdec_hevc.c:49
ret
ret
Definition: filter_design.txt:187
AVStream
Stream structure.
Definition: avformat.h:865
pos
unsigned int pos
Definition: spdifenc.c:412
avformat.h
PayloadContext::vps
uint8_t * vps
Definition: rtpdec_hevc.c:43
AVStream::index
int index
stream index in AVFormatContext
Definition: avformat.h:866
pkt
static AVPacket pkt
Definition: demuxing_decoding.c:54
AVPacket::stream_index
int stream_index
Definition: packet.h:357
AVMEDIA_TYPE_VIDEO
@ AVMEDIA_TYPE_VIDEO
Definition: avutil.h:201
ff_h264_handle_aggregated_packet
int ff_h264_handle_aggregated_packet(AVFormatContext *ctx, PayloadContext *data, AVPacket *pkt, const uint8_t *buf, int len, int start_skip, int *nal_counters, int nal_mask)
Definition: rtpdec_h264.c:206
PayloadContext::sei_size
int sei_size
Definition: rtpdec_hevc.c:44
AVPacket
This structure stores compressed data.
Definition: packet.h:332
av_freep
#define av_freep(p)
Definition: tableprint_vlc.h:35
AVSTREAM_PARSE_FULL
@ AVSTREAM_PARSE_FULL
full parsing and repack
Definition: avformat.h:786
flags
#define flags(name, subs,...)
Definition: cbs_av1.c:565
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
avstring.h
PayloadContext
RTP/JPEG specific private data.
Definition: rdt.c:83
PayloadContext::pps_size
int pps_size
Definition: rtpdec_hevc.c:44
ff_h264_parse_sprop_parameter_sets
int ff_h264_parse_sprop_parameter_sets(AVFormatContext *s, uint8_t **data_ptr, int *size_ptr, const char *value)
Definition: rtpdec_h264.c:96
line
The official guide to swscale for confused that consecutive non overlapping rectangles of slice_bottom special converter These generally are unscaled converters of common like for each output line the vertical scaler pulls lines from a ring buffer When the ring buffer does not contain the wanted line
Definition: swscale.txt:40
RTPDynamicProtocolHandler
Definition: rtpdec.h:115
ff_alloc_extradata
int ff_alloc_extradata(AVCodecParameters *par, int size)
Allocate extradata with additional AV_INPUT_BUFFER_PADDING_SIZE at end which is always set to 0.
Definition: utils.c:3328