1 /*
2 * Common code for the RTP depacketization of MPEG-4 formats.
3 * Copyright (c) 2010 Fabrice Bellard
4 * Romain Degez
5 *
6 * This file is part of FFmpeg.
7 *
8 * FFmpeg is free software; you can redistribute it and/or
9 * modify it under the terms of the GNU Lesser General Public
10 * License as published by the Free Software Foundation; either
11 * version 2.1 of the License, or (at your option) any later version.
12 *
13 * FFmpeg is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
16 * Lesser General Public License for more details.
17 *
18 * You should have received a copy of the GNU Lesser General Public
19 * License along with FFmpeg; if not, write to the Free Software
20 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
21 */
22
23 /**
24 * @file
25 * @brief MPEG-4 / RTP Code
26 * @author Fabrice Bellard
27 * @author Romain Degez
28 */
29
30 #include "rtpdec_formats.h"
31 #include "internal.h"
32 #include "libavutil/attributes.h"
33 #include "libavutil/avstring.h"
34 #include "libavcodec/get_bits.h"
35
36 #define MAX_AAC_HBR_FRAME_SIZE 8191
37
38 /** Structure listing useful vars to parse RTP packet payload */
39 struct PayloadContext {
40 int sizelength;
41 int indexlength;
42 int indexdeltalength;
43 int profile_level_id;
44 int streamtype;
45 int objecttype;
46 char *mode;
47
48 /** mpeg 4 AU headers */
49 struct AUHeaders {
50 int size;
51 int index;
52 int cts_flag;
53 int cts;
54 int dts_flag;
55 int dts;
56 int rap_flag;
57 int streamstate;
58 } *au_headers;
59 int au_headers_allocated;
60 int nb_au_headers;
61 int au_headers_length_bytes;
62 int cur_au_index;
63
64 uint8_t buf[FFMAX(RTP_MAX_PACKET_LENGTH, MAX_AAC_HBR_FRAME_SIZE)];
65 int buf_pos, buf_size;
66 uint32_t timestamp;
67 };
68
69 typedef struct AttrNameMap {
70 const char *str;
71 uint16_t type;
72 uint32_t offset;
73
74 /** Range for integer values */
75 struct Range {
76 int min;
77 int max;
78 } range;
79 } AttrNameMap;
80
81 /* All known fmtp parameters and the corresponding RTPAttrTypeEnum */
82 #define ATTR_NAME_TYPE_INT 0
83 #define ATTR_NAME_TYPE_STR 1
84 static const AttrNameMap attr_names[] = {
85 { "SizeLength", ATTR_NAME_TYPE_INT,
86 offsetof(PayloadContext, sizelength),
87 {0, 32} }, // SizeLength number of bits used to encode AU-size integer value
88 { "IndexLength", ATTR_NAME_TYPE_INT,
89 offsetof(PayloadContext, indexlength),
90 {0, 32} }, // IndexLength number of bits used to encode AU-Index integer value
91 { "IndexDeltaLength", ATTR_NAME_TYPE_INT,
92 offsetof(PayloadContext, indexdeltalength),
93 {0, 32} }, // IndexDeltaLength number of bits to encode AU-Index-delta integer value
94 { "profile-level-id", ATTR_NAME_TYPE_INT,
95 offsetof(PayloadContext, profile_level_id),
96 {INT32_MIN, INT32_MAX} }, // It differs depending on StreamType
97 { "StreamType", ATTR_NAME_TYPE_INT,
98 offsetof(PayloadContext, streamtype),
99 {0x00, 0x3F} }, // Values from ISO/IEC 14496-1, 'StreamType Values' table
100 { "mode", ATTR_NAME_TYPE_STR,
101 offsetof(PayloadContext, mode),
102 {0} },
103 { NULL, -1, -1, {0} },
104 };
105
close_context(PayloadContext * data)106 static void close_context(PayloadContext *data)
107 {
108 av_freep(&data->au_headers);
109 av_freep(&data->mode);
110 }
111
parse_fmtp_config(AVCodecParameters * par,const char * value)112 static int parse_fmtp_config(AVCodecParameters *par, const char *value)
113 {
114 /* decode the hexa encoded parameter */
115 int len = ff_hex_to_data(NULL, value), ret;
116
117 if ((ret = ff_alloc_extradata(par, len)) < 0)
118 return ret;
119 ff_hex_to_data(par->extradata, value);
120 return 0;
121 }
122
rtp_parse_mp4_au(PayloadContext * data,const uint8_t * buf,int len)123 static int rtp_parse_mp4_au(PayloadContext *data, const uint8_t *buf, int len)
124 {
125 int au_headers_length, au_header_size, i;
126 GetBitContext getbitcontext;
127 int ret;
128
129 if (len < 2)
130 return AVERROR_INVALIDDATA;
131
132 /* decode the first 2 bytes where the AUHeader sections are stored
133 length in bits */
134 au_headers_length = AV_RB16(buf);
135
136 if (au_headers_length > RTP_MAX_PACKET_LENGTH)
137 return -1;
138
139 data->au_headers_length_bytes = (au_headers_length + 7) / 8;
140
141 /* skip AU headers length section (2 bytes) */
142 buf += 2;
143 len -= 2;
144
145 if (len < data->au_headers_length_bytes)
146 return AVERROR_INVALIDDATA;
147
148 ret = init_get_bits(&getbitcontext, buf, data->au_headers_length_bytes * 8);
149 if (ret < 0)
150 return ret;
151
152 /* XXX: Wrong if optional additional sections are present (cts, dts etc...) */
153 au_header_size = data->sizelength + data->indexlength;
154 if (au_header_size <= 0 || (au_headers_length % au_header_size != 0))
155 return -1;
156
157 data->nb_au_headers = au_headers_length / au_header_size;
158 if (!data->au_headers || data->au_headers_allocated < data->nb_au_headers) {
159 av_free(data->au_headers);
160 data->au_headers = av_malloc(sizeof(struct AUHeaders) * data->nb_au_headers);
161 if (!data->au_headers)
162 return AVERROR(ENOMEM);
163 data->au_headers_allocated = data->nb_au_headers;
164 }
165
166 for (i = 0; i < data->nb_au_headers; ++i) {
167 data->au_headers[i].size = get_bits_long(&getbitcontext, data->sizelength);
168 data->au_headers[i].index = get_bits_long(&getbitcontext, data->indexlength);
169 }
170
171 return 0;
172 }
173
174
175 /* Follows RFC 3640 */
aac_parse_packet(AVFormatContext * ctx,PayloadContext * data,AVStream * st,AVPacket * pkt,uint32_t * timestamp,const uint8_t * buf,int len,uint16_t seq,int flags)176 static int aac_parse_packet(AVFormatContext *ctx, PayloadContext *data,
177 AVStream *st, AVPacket *pkt, uint32_t *timestamp,
178 const uint8_t *buf, int len, uint16_t seq,
179 int flags)
180 {
181 int ret;
182
183
184 if (!buf) {
185 if (data->cur_au_index > data->nb_au_headers) {
186 av_log(ctx, AV_LOG_ERROR, "Invalid parser state\n");
187 return AVERROR_INVALIDDATA;
188 }
189 if (data->buf_size - data->buf_pos < data->au_headers[data->cur_au_index].size) {
190 av_log(ctx, AV_LOG_ERROR, "Invalid AU size\n");
191 return AVERROR_INVALIDDATA;
192 }
193 if ((ret = av_new_packet(pkt, data->au_headers[data->cur_au_index].size)) < 0) {
194 av_log(ctx, AV_LOG_ERROR, "Out of memory\n");
195 return ret;
196 }
197 memcpy(pkt->data, &data->buf[data->buf_pos], data->au_headers[data->cur_au_index].size);
198 data->buf_pos += data->au_headers[data->cur_au_index].size;
199 pkt->stream_index = st->index;
200 data->cur_au_index++;
201
202 if (data->cur_au_index == data->nb_au_headers) {
203 data->buf_pos = 0;
204 return 0;
205 }
206
207 return 1;
208 }
209
210 if (rtp_parse_mp4_au(data, buf, len)) {
211 av_log(ctx, AV_LOG_ERROR, "Error parsing AU headers\n");
212 return -1;
213 }
214
215 buf += data->au_headers_length_bytes + 2;
216 len -= data->au_headers_length_bytes + 2;
217 if (data->nb_au_headers == 1 && len < data->au_headers[0].size) {
218 /* Packet is fragmented */
219
220 if (!data->buf_pos) {
221 if (data->au_headers[0].size > MAX_AAC_HBR_FRAME_SIZE) {
222 av_log(ctx, AV_LOG_ERROR, "Invalid AU size\n");
223 return AVERROR_INVALIDDATA;
224 }
225
226 data->buf_size = data->au_headers[0].size;
227 data->timestamp = *timestamp;
228 }
229
230 if (data->timestamp != *timestamp ||
231 data->au_headers[0].size != data->buf_size ||
232 data->buf_pos + len > MAX_AAC_HBR_FRAME_SIZE) {
233 data->buf_pos = 0;
234 data->buf_size = 0;
235 av_log(ctx, AV_LOG_ERROR, "Invalid packet received\n");
236 return AVERROR_INVALIDDATA;
237 }
238
239 memcpy(&data->buf[data->buf_pos], buf, len);
240 data->buf_pos += len;
241
242 if (!(flags & RTP_FLAG_MARKER))
243 return AVERROR(EAGAIN);
244
245 if (data->buf_pos != data->buf_size) {
246 data->buf_pos = 0;
247 av_log(ctx, AV_LOG_ERROR, "Missed some packets, discarding frame\n");
248 return AVERROR_INVALIDDATA;
249 }
250
251 data->buf_pos = 0;
252 ret = av_new_packet(pkt, data->buf_size);
253 if (ret < 0) {
254 av_log(ctx, AV_LOG_ERROR, "Out of memory\n");
255 return ret;
256 }
257 pkt->stream_index = st->index;
258
259 memcpy(pkt->data, data->buf, data->buf_size);
260
261 return 0;
262 }
263
264 if (len < data->au_headers[0].size) {
265 av_log(ctx, AV_LOG_ERROR, "First AU larger than packet size\n");
266 return AVERROR_INVALIDDATA;
267 }
268 if ((ret = av_new_packet(pkt, data->au_headers[0].size)) < 0) {
269 av_log(ctx, AV_LOG_ERROR, "Out of memory\n");
270 return ret;
271 }
272 memcpy(pkt->data, buf, data->au_headers[0].size);
273 len -= data->au_headers[0].size;
274 buf += data->au_headers[0].size;
275 pkt->stream_index = st->index;
276
277 if (len > 0 && data->nb_au_headers > 1) {
278 data->buf_size = FFMIN(len, sizeof(data->buf));
279 memcpy(data->buf, buf, data->buf_size);
280 data->cur_au_index = 1;
281 data->buf_pos = 0;
282 return 1;
283 }
284
285 return 0;
286 }
287
parse_fmtp(AVFormatContext * s,AVStream * stream,PayloadContext * data,const char * attr,const char * value)288 static int parse_fmtp(AVFormatContext *s,
289 AVStream *stream, PayloadContext *data,
290 const char *attr, const char *value)
291 {
292 AVCodecParameters *par = stream->codecpar;
293 int res, i;
294
295 if (!strcmp(attr, "config")) {
296 res = parse_fmtp_config(par, value);
297
298 if (res < 0)
299 return res;
300 }
301
302 if (par->codec_id == AV_CODEC_ID_AAC) {
303 /* Looking for a known attribute */
304 for (i = 0; attr_names[i].str; ++i) {
305 if (!av_strcasecmp(attr, attr_names[i].str)) {
306 if (attr_names[i].type == ATTR_NAME_TYPE_INT) {
307 char *end_ptr = NULL;
308 long long int val = strtoll(value, &end_ptr, 10);
309 if (end_ptr == value || end_ptr[0] != '\0') {
310 av_log(s, AV_LOG_ERROR,
311 "The %s field value is not a valid number: %s\n",
312 attr, value);
313 return AVERROR_INVALIDDATA;
314 }
315 if (val < attr_names[i].range.min ||
316 val > attr_names[i].range.max) {
317 av_log(s, AV_LOG_ERROR,
318 "fmtp field %s should be in range [%d,%d] (provided value: %lld)",
319 attr, attr_names[i].range.min, attr_names[i].range.max, val);
320 return AVERROR_INVALIDDATA;
321 }
322
323 *(int *)((char *)data+
324 attr_names[i].offset) = (int) val;
325 } else if (attr_names[i].type == ATTR_NAME_TYPE_STR) {
326 char *val = av_strdup(value);
327 if (!val)
328 return AVERROR(ENOMEM);
329 *(char **)((char *)data+
330 attr_names[i].offset) = val;
331 }
332 }
333 }
334 }
335 return 0;
336 }
337
parse_sdp_line(AVFormatContext * s,int st_index,PayloadContext * data,const char * line)338 static int parse_sdp_line(AVFormatContext *s, int st_index,
339 PayloadContext *data, const char *line)
340 {
341 const char *p;
342
343 if (st_index < 0)
344 return 0;
345
346 if (av_strstart(line, "fmtp:", &p))
347 return ff_parse_fmtp(s, s->streams[st_index], data, p, parse_fmtp);
348
349 return 0;
350 }
351
352 const RTPDynamicProtocolHandler ff_mp4v_es_dynamic_handler = {
353 .enc_name = "MP4V-ES",
354 .codec_type = AVMEDIA_TYPE_VIDEO,
355 .codec_id = AV_CODEC_ID_MPEG4,
356 .need_parsing = AVSTREAM_PARSE_FULL,
357 .priv_data_size = sizeof(PayloadContext),
358 .parse_sdp_a_line = parse_sdp_line,
359 };
360
361 const RTPDynamicProtocolHandler ff_mpeg4_generic_dynamic_handler = {
362 .enc_name = "mpeg4-generic",
363 .codec_type = AVMEDIA_TYPE_AUDIO,
364 .codec_id = AV_CODEC_ID_AAC,
365 .priv_data_size = sizeof(PayloadContext),
366 .parse_sdp_a_line = parse_sdp_line,
367 .close = close_context,
368 .parse_packet = aac_parse_packet,
369 };
370