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 "rtpdec.h"
29 #include "rtpdec_formats.h"
30 
31 #define RTP_HEVC_PAYLOAD_HEADER_SIZE 2
32 #define RTP_HEVC_FU_HEADER_SIZE 1
33 #define RTP_HEVC_DONL_FIELD_SIZE 2
34 #define RTP_HEVC_DOND_FIELD_SIZE 1
35 #define RTP_HEVC_AP_NALU_LENGTH_FIELD_SIZE 2
36 #define HEVC_SPECIFIED_NAL_UNIT_TYPES 48
37 
38 /* SDP out-of-band signaling data */
39 struct PayloadContext {
42  uint8_t *sps, *pps, *vps, *sei;
44 };
45 
46 static const uint8_t start_sequence[] = { 0x00, 0x00, 0x00, 0x01 };
47 
49  AVStream *stream,
50  PayloadContext *hevc_data,
51  const char *attr, const char *value)
52 {
53  /* profile-space: 0-3 */
54  /* profile-id: 0-31 */
55  if (!strcmp(attr, "profile-id")) {
56  hevc_data->profile_id = atoi(value);
57  av_log(s, AV_LOG_TRACE, "SDP: found profile-id: %d\n", hevc_data->profile_id);
58  }
59 
60  /* tier-flag: 0-1 */
61  /* level-id: 0-255 */
62  /* interop-constraints: [base16] */
63  /* profile-compatibility-indicator: [base16] */
64  /* sprop-sub-layer-id: 0-6, defines highest possible value for TID, default: 6 */
65  /* recv-sub-layer-id: 0-6 */
66  /* max-recv-level-id: 0-255 */
67  /* tx-mode: MSM,SSM */
68  /* sprop-vps: [base64] */
69  /* sprop-sps: [base64] */
70  /* sprop-pps: [base64] */
71  /* sprop-sei: [base64] */
72  if (!strcmp(attr, "sprop-vps") || !strcmp(attr, "sprop-sps") ||
73  !strcmp(attr, "sprop-pps") || !strcmp(attr, "sprop-sei")) {
74  uint8_t **data_ptr = NULL;
75  int *size_ptr = NULL;
76  if (!strcmp(attr, "sprop-vps")) {
77  data_ptr = &hevc_data->vps;
78  size_ptr = &hevc_data->vps_size;
79  } else if (!strcmp(attr, "sprop-sps")) {
80  data_ptr = &hevc_data->sps;
81  size_ptr = &hevc_data->sps_size;
82  } else if (!strcmp(attr, "sprop-pps")) {
83  data_ptr = &hevc_data->pps;
84  size_ptr = &hevc_data->pps_size;
85  } else if (!strcmp(attr, "sprop-sei")) {
86  data_ptr = &hevc_data->sei;
87  size_ptr = &hevc_data->sei_size;
88  } else
89  av_assert0(0);
90 
92  size_ptr, value);
93  }
94 
95  /* max-lsr, max-lps, max-cpb, max-dpb, max-br, max-tr, max-tc */
96  /* max-fps */
97 
98  /* sprop-max-don-diff: 0-32767
99 
100  When the RTP stream depends on one or more other RTP
101  streams (in this case tx-mode MUST be equal to "MSM" and
102  MSM is in use), this parameter MUST be present and the
103  value MUST be greater than 0.
104  */
105  if (!strcmp(attr, "sprop-max-don-diff")) {
106  if (atoi(value) > 0)
107  hevc_data->using_donl_field = 1;
108  av_log(s, AV_LOG_TRACE, "Found sprop-max-don-diff in SDP, DON field usage is: %d\n",
109  hevc_data->using_donl_field);
110  }
111 
112  /* sprop-depack-buf-nalus: 0-32767 */
113  if (!strcmp(attr, "sprop-depack-buf-nalus")) {
114  if (atoi(value) > 0)
115  hevc_data->using_donl_field = 1;
116  av_log(s, AV_LOG_TRACE, "Found sprop-depack-buf-nalus in SDP, DON field usage is: %d\n",
117  hevc_data->using_donl_field);
118  }
119 
120  /* sprop-depack-buf-bytes: 0-4294967295 */
121  /* depack-buf-cap */
122  /* sprop-segmentation-id: 0-3 */
123  /* sprop-spatial-segmentation-idc: [base16] */
124  /* dec-parallel-ca: */
125  /* include-dph */
126 
127  return 0;
128 }
129 
130 static av_cold int hevc_parse_sdp_line(AVFormatContext *ctx, int st_index,
131  PayloadContext *hevc_data, const char *line)
132 {
133  AVStream *current_stream;
134  AVCodecParameters *par;
135  const char *sdp_line_ptr = line;
136 
137  if (st_index < 0)
138  return 0;
139 
140  current_stream = ctx->streams[st_index];
141  par = current_stream->codecpar;
142 
143  if (av_strstart(sdp_line_ptr, "framesize:", &sdp_line_ptr)) {
144  ff_h264_parse_framesize(par, sdp_line_ptr);
145  } else if (av_strstart(sdp_line_ptr, "fmtp:", &sdp_line_ptr)) {
146  int ret = ff_parse_fmtp(ctx, current_stream, hevc_data, sdp_line_ptr,
148  if (hevc_data->vps_size || hevc_data->sps_size ||
149  hevc_data->pps_size || hevc_data->sei_size) {
150  av_freep(&par->extradata);
151  par->extradata_size = hevc_data->vps_size + hevc_data->sps_size +
152  hevc_data->pps_size + hevc_data->sei_size;
153  par->extradata = av_malloc(par->extradata_size +
155  if (!par->extradata) {
156  ret = AVERROR(ENOMEM);
157  par->extradata_size = 0;
158  } else {
159  int pos = 0;
160  memcpy(par->extradata + pos, hevc_data->vps, hevc_data->vps_size);
161  pos += hevc_data->vps_size;
162  memcpy(par->extradata + pos, hevc_data->sps, hevc_data->sps_size);
163  pos += hevc_data->sps_size;
164  memcpy(par->extradata + pos, hevc_data->pps, hevc_data->pps_size);
165  pos += hevc_data->pps_size;
166  memcpy(par->extradata + pos, hevc_data->sei, hevc_data->sei_size);
167  pos += hevc_data->sei_size;
168  memset(par->extradata + pos, 0, AV_INPUT_BUFFER_PADDING_SIZE);
169  }
170 
171  av_freep(&hevc_data->vps);
172  av_freep(&hevc_data->sps);
173  av_freep(&hevc_data->pps);
174  av_freep(&hevc_data->sei);
175  hevc_data->vps_size = 0;
176  hevc_data->sps_size = 0;
177  hevc_data->pps_size = 0;
178  hevc_data->sei_size = 0;
179  }
180  return ret;
181  }
182 
183  return 0;
184 }
185 
187  AVStream *st, AVPacket *pkt, uint32_t *timestamp,
188  const uint8_t *buf, int len, uint16_t seq,
189  int flags)
190 {
191  const uint8_t *rtp_pl = buf;
192  int tid, lid, nal_type;
193  int first_fragment, last_fragment, fu_type;
194  uint8_t new_nal_header[2];
195  int res = 0;
196 
197  /* sanity check for size of input packet: 1 byte payload at least */
198  if (len < RTP_HEVC_PAYLOAD_HEADER_SIZE + 1) {
199  av_log(ctx, AV_LOG_ERROR, "Too short RTP/HEVC packet, got %d bytes\n", len);
200  return AVERROR_INVALIDDATA;
201  }
202 
203  /*
204  * decode the HEVC payload header according to section 4 of draft version 6:
205  *
206  * 0 1
207  * 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
208  * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
209  * |F| Type | LayerId | TID |
210  * +-------------+-----------------+
211  *
212  * Forbidden zero (F): 1 bit
213  * NAL unit type (Type): 6 bits
214  * NUH layer ID (LayerId): 6 bits
215  * NUH temporal ID plus 1 (TID): 3 bits
216  */
217  nal_type = (buf[0] >> 1) & 0x3f;
218  lid = ((buf[0] << 5) & 0x20) | ((buf[1] >> 3) & 0x1f);
219  tid = buf[1] & 0x07;
220 
221  /* sanity check for correct layer ID */
222  if (lid) {
223  /* future scalable or 3D video coding extensions */
224  avpriv_report_missing_feature(ctx, "Multi-layer HEVC coding");
225  return AVERROR_PATCHWELCOME;
226  }
227 
228  /* sanity check for correct temporal ID */
229  if (!tid) {
230  av_log(ctx, AV_LOG_ERROR, "Illegal temporal ID in RTP/HEVC packet\n");
231  return AVERROR_INVALIDDATA;
232  }
233 
234  /* sanity check for correct NAL unit type */
235  if (nal_type > 50) {
236  av_log(ctx, AV_LOG_ERROR, "Unsupported (HEVC) NAL type (%d)\n", nal_type);
237  return AVERROR_INVALIDDATA;
238  }
239 
240  switch (nal_type) {
241  /* video parameter set (VPS) */
242  case 32:
243  /* sequence parameter set (SPS) */
244  case 33:
245  /* picture parameter set (PPS) */
246  case 34:
247  /* supplemental enhancement information (SEI) */
248  case 39:
249  /* single NAL unit packet */
250  default:
251  /* create A/V packet */
252  if ((res = av_new_packet(pkt, sizeof(start_sequence) + len)) < 0)
253  return res;
254  /* A/V packet: copy start sequence */
255  memcpy(pkt->data, start_sequence, sizeof(start_sequence));
256  /* A/V packet: copy NAL unit data */
257  memcpy(pkt->data + sizeof(start_sequence), buf, len);
258 
259  break;
260  /* aggregated packet (AP) - with two or more NAL units */
261  case 48:
262  /* pass the HEVC payload header */
265 
266  /* pass the HEVC DONL field */
267  if (rtp_hevc_ctx->using_donl_field) {
270  }
271 
272  res = ff_h264_handle_aggregated_packet(ctx, rtp_hevc_ctx, pkt, buf, len,
273  rtp_hevc_ctx->using_donl_field ?
275  NULL, 0);
276  if (res < 0)
277  return res;
278  break;
279  /* fragmentation unit (FU) */
280  case 49:
281  /* pass the HEVC payload header */
284 
285  /*
286  * decode the FU header
287  *
288  * 0 1 2 3 4 5 6 7
289  * +-+-+-+-+-+-+-+-+
290  * |S|E| FuType |
291  * +---------------+
292  *
293  * Start fragment (S): 1 bit
294  * End fragment (E): 1 bit
295  * FuType: 6 bits
296  */
297  first_fragment = buf[0] & 0x80;
298  last_fragment = buf[0] & 0x40;
299  fu_type = buf[0] & 0x3f;
300 
301  /* pass the HEVC FU header */
304 
305  /* pass the HEVC DONL field */
306  if (rtp_hevc_ctx->using_donl_field) {
309  }
310 
311  av_log(ctx, AV_LOG_TRACE, " FU type %d with %d bytes\n", fu_type, len);
312 
313  /* sanity check for size of input packet: 1 byte payload at least */
314  if (len <= 0) {
315  if (len < 0) {
316  av_log(ctx, AV_LOG_ERROR,
317  "Too short RTP/HEVC packet, got %d bytes of NAL unit type %d\n",
318  len, nal_type);
319  return AVERROR_INVALIDDATA;
320  } else {
321  return AVERROR(EAGAIN);
322  }
323  }
324 
325  if (first_fragment && last_fragment) {
326  av_log(ctx, AV_LOG_ERROR, "Illegal combination of S and E bit in RTP/HEVC packet\n");
327  return AVERROR_INVALIDDATA;
328  }
329 
330  new_nal_header[0] = (rtp_pl[0] & 0x81) | (fu_type << 1);
331  new_nal_header[1] = rtp_pl[1];
332 
333  res = ff_h264_handle_frag_packet(pkt, buf, len, first_fragment,
334  new_nal_header, sizeof(new_nal_header));
335 
336  break;
337  /* PACI packet */
338  case 50:
339  /* Temporal scalability control information (TSCI) */
340  avpriv_report_missing_feature(ctx, "PACI packets for RTP/HEVC");
341  res = AVERROR_PATCHWELCOME;
342  break;
343  }
344 
345  pkt->stream_index = st->index;
346 
347  return res;
348 }
349 
351  .enc_name = "H265",
352  .codec_type = AVMEDIA_TYPE_VIDEO,
353  .codec_id = AV_CODEC_ID_HEVC,
354  .need_parsing = AVSTREAM_PARSE_FULL,
355  .priv_data_size = sizeof(PayloadContext),
356  .parse_sdp_a_line = hevc_parse_sdp_line,
358 };
AVPacket pkt
Definition: rtpdec_qt.c:37
#define RTP_HEVC_DOND_FIELD_SIZE
Definition: rtpdec_hevc.c:34
#define NULL
Definition: coverity.c:32
#define AVERROR_INVALIDDATA
Invalid data found when processing input.
Definition: error.h:59
int using_donl_field
Definition: rtpdec_hevc.c:40
static const uint8_t start_sequence[]
Definition: rtpdec_hevc.c:46
RTP/JPEG specific private data.
Definition: rdt.c:83
int index
stream index in AVFormatContext
Definition: avformat.h:882
const RTPDynamicProtocolHandler ff_hevc_dynamic_handler
Definition: rtpdec_hevc.c:350
static av_cold int hevc_parse_sdp_line(AVFormatContext *ctx, int st_index, PayloadContext *hevc_data, const char *line)
Definition: rtpdec_hevc.c:130
This struct describes the properties of an encoded stream.
Definition: avcodec.h:3945
uint8_t * sei
Definition: rtpdec_hevc.c:42
Format I/O context.
Definition: avformat.h:1358
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:48
#define av_assert0(cond)
assert() equivalent, that is always enabled.
Definition: avassert.h:37
uint8_t
#define av_cold
Definition: attributes.h:82
#define av_malloc(s)
uint8_t * pps
Definition: rtpdec_hevc.c:42
#define AV_LOG_TRACE
Extremely verbose debugging, useful for libav* development.
Definition: log.h:202
GLsizei GLboolean const GLfloat * value
Definition: opengl_enc.c:108
AVStream ** streams
A list of all streams in the file.
Definition: avformat.h:1426
uint8_t * data
Definition: avcodec.h:1477
bitstream reader API header.
#define av_log(a,...)
int av_new_packet(AVPacket *pkt, int size)
Allocate the payload of a packet and initialize its fields with default values.
Definition: avpacket.c:86
uint32_t timestamp
current frame timestamp
Definition: rtpdec_ac3.c:31
#define RTP_HEVC_FU_HEADER_SIZE
Definition: rtpdec_hevc.c:32
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
Definition: log.h:176
Definition: graph2dot.c:48
simple assert() macros that are a bit more flexible than ISO C assert().
uint8_t * sps
Definition: rtpdec_hevc.c:42
int extradata_size
Size of the extradata content in bytes.
Definition: avcodec.h:3971
uint8_t * vps
Definition: rtpdec_hevc.c:42
AVFormatContext * ctx
Definition: movenc.c:48
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:186
#define s(width, name)
Definition: cbs_vp9.c:257
Stream structure.
Definition: avformat.h:881
#define AVERROR_PATCHWELCOME
Not yet implemented in FFmpeg, patches welcome.
Definition: error.h:62
void ff_h264_parse_framesize(AVCodecParameters *par, const char *p)
Definition: rtpdec_h264.c:184
#define RTP_HEVC_PAYLOAD_HEADER_SIZE
Definition: rtpdec_hevc.c:31
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
uint8_t * buf
the temporary storage buffer
Definition: rtpdec_asf.c:183
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:889
const char * enc_name
Definition: rtpdec.h:116
int ff_h264_parse_sprop_parameter_sets(AVFormatContext *s, uint8_t **data_ptr, int *size_ptr, const char *value)
Definition: rtpdec_h264.c:96
void avpriv_report_missing_feature(void *avc, const char *msg,...) av_printf_format(2
Log a generic warning message about a missing feature.
#define flags(name, subs,...)
Definition: cbs_av1.c:561
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
full parsing and repack
Definition: avformat.h:800
Main libavformat public API header.
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:33
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
#define AV_INPUT_BUFFER_PADDING_SIZE
Required number of additionally allocated bytes at the end of the input bitstream for decoding...
Definition: avcodec.h:790
static int parse_packet(AVFormatContext *s, AVPacket *pkt, int stream_index)
Parse a packet, add all split parts to parse_queue.
Definition: utils.c:1438
uint8_t * extradata
Extra binary data needed for initializing the decoder, codec-dependent.
Definition: avcodec.h:3967
#define RTP_HEVC_DONL_FIELD_SIZE
Definition: rtpdec_hevc.c:33
#define av_freep(p)
AVCodecParameters * codecpar
Codec parameters associated with this stream.
Definition: avformat.h:1028
int stream_index
Definition: avcodec.h:1479
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
This structure stores compressed data.
Definition: avcodec.h:1454