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
128 if (len < 2)
129 return AVERROR_INVALIDDATA;
130
131 /* decode the first 2 bytes where the AUHeader sections are stored
132 length in bits */
133 au_headers_length = AV_RB16(buf);
134
135 if (au_headers_length > RTP_MAX_PACKET_LENGTH)
136 return -1;
137
138 data->au_headers_length_bytes = (au_headers_length + 7) / 8;
139
140 /* skip AU headers length section (2 bytes) */
141 buf += 2;
142 len -= 2;
143
144 if (len < data->au_headers_length_bytes)
145 return AVERROR_INVALIDDATA;
146
147 init_get_bits(&getbitcontext, buf, data->au_headers_length_bytes * 8);
148
149 /* XXX: Wrong if optional additional sections are present (cts, dts etc...) */
150 au_header_size = data->sizelength + data->indexlength;
151 if (au_header_size <= 0 || (au_headers_length % au_header_size != 0))
152 return -1;
153
154 data->nb_au_headers = au_headers_length / au_header_size;
155 if (!data->au_headers || data->au_headers_allocated < data->nb_au_headers) {
156 av_free(data->au_headers);
157 data->au_headers = av_malloc(sizeof(struct AUHeaders) * data->nb_au_headers);
158 if (!data->au_headers)
159 return AVERROR(ENOMEM);
160 data->au_headers_allocated = data->nb_au_headers;
161 }
162
163 for (i = 0; i < data->nb_au_headers; ++i) {
164 data->au_headers[i].size = get_bits_long(&getbitcontext, data->sizelength);
165 data->au_headers[i].index = get_bits_long(&getbitcontext, data->indexlength);
166 }
167
168 return 0;
169 }
170
171
172 /* 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)173 static int aac_parse_packet(AVFormatContext *ctx, PayloadContext *data,
174 AVStream *st, AVPacket *pkt, uint32_t *timestamp,
175 const uint8_t *buf, int len, uint16_t seq,
176 int flags)
177 {
178 int ret;
179
180
181 if (!buf) {
182 if (data->cur_au_index > data->nb_au_headers) {
183 av_log(ctx, AV_LOG_ERROR, "Invalid parser state\n");
184 return AVERROR_INVALIDDATA;
185 }
186 if (data->buf_size - data->buf_pos < data->au_headers[data->cur_au_index].size) {
187 av_log(ctx, AV_LOG_ERROR, "Invalid AU size\n");
188 return AVERROR_INVALIDDATA;
189 }
190 if ((ret = av_new_packet(pkt, data->au_headers[data->cur_au_index].size)) < 0) {
191 av_log(ctx, AV_LOG_ERROR, "Out of memory\n");
192 return ret;
193 }
194 memcpy(pkt->data, &data->buf[data->buf_pos], data->au_headers[data->cur_au_index].size);
195 data->buf_pos += data->au_headers[data->cur_au_index].size;
196 pkt->stream_index = st->index;
197 data->cur_au_index++;
198
199 if (data->cur_au_index == data->nb_au_headers) {
200 data->buf_pos = 0;
201 return 0;
202 }
203
204 return 1;
205 }
206
207 if (rtp_parse_mp4_au(data, buf, len)) {
208 av_log(ctx, AV_LOG_ERROR, "Error parsing AU headers\n");
209 return -1;
210 }
211
212 buf += data->au_headers_length_bytes + 2;
213 len -= data->au_headers_length_bytes + 2;
214 if (data->nb_au_headers == 1 && len < data->au_headers[0].size) {
215 /* Packet is fragmented */
216
217 if (!data->buf_pos) {
218 if (data->au_headers[0].size > MAX_AAC_HBR_FRAME_SIZE) {
219 av_log(ctx, AV_LOG_ERROR, "Invalid AU size\n");
220 return AVERROR_INVALIDDATA;
221 }
222
223 data->buf_size = data->au_headers[0].size;
224 data->timestamp = *timestamp;
225 }
226
227 if (data->timestamp != *timestamp ||
228 data->au_headers[0].size != data->buf_size ||
229 data->buf_pos + len > MAX_AAC_HBR_FRAME_SIZE) {
230 data->buf_pos = 0;
231 data->buf_size = 0;
232 av_log(ctx, AV_LOG_ERROR, "Invalid packet received\n");
233 return AVERROR_INVALIDDATA;
234 }
235
236 memcpy(&data->buf[data->buf_pos], buf, len);
237 data->buf_pos += len;
238
239 if (!(flags & RTP_FLAG_MARKER))
240 return AVERROR(EAGAIN);
241
242 if (data->buf_pos != data->buf_size) {
243 data->buf_pos = 0;
244 av_log(ctx, AV_LOG_ERROR, "Missed some packets, discarding frame\n");
245 return AVERROR_INVALIDDATA;
246 }
247
248 data->buf_pos = 0;
249 ret = av_new_packet(pkt, data->buf_size);
250 if (ret < 0) {
251 av_log(ctx, AV_LOG_ERROR, "Out of memory\n");
252 return ret;
253 }
254 pkt->stream_index = st->index;
255
256 memcpy(pkt->data, data->buf, data->buf_size);
257
258 return 0;
259 }
260
261 if (len < data->au_headers[0].size) {
262 av_log(ctx, AV_LOG_ERROR, "First AU larger than packet size\n");
263 return AVERROR_INVALIDDATA;
264 }
265 if ((ret = av_new_packet(pkt, data->au_headers[0].size)) < 0) {
266 av_log(ctx, AV_LOG_ERROR, "Out of memory\n");
267 return ret;
268 }
269 memcpy(pkt->data, buf, data->au_headers[0].size);
270 len -= data->au_headers[0].size;
271 buf += data->au_headers[0].size;
272 pkt->stream_index = st->index;
273
274 if (len > 0 && data->nb_au_headers > 1) {
275 data->buf_size = FFMIN(len, sizeof(data->buf));
276 memcpy(data->buf, buf, data->buf_size);
277 data->cur_au_index = 1;
278 data->buf_pos = 0;
279 return 1;
280 }
281
282 return 0;
283 }
284
parse_fmtp(AVFormatContext * s,AVStream * stream,PayloadContext * data,const char * attr,const char * value)285 static int parse_fmtp(AVFormatContext *s,
286 AVStream *stream, PayloadContext *data,
287 const char *attr, const char *value)
288 {
289 AVCodecParameters *par = stream->codecpar;
290 int res, i;
291
292 if (!strcmp(attr, "config")) {
293 res = parse_fmtp_config(par, value);
294
295 if (res < 0)
296 return res;
297 }
298
299 if (par->codec_id == AV_CODEC_ID_AAC) {
300 /* Looking for a known attribute */
301 for (i = 0; attr_names[i].str; ++i) {
302 if (!av_strcasecmp(attr, attr_names[i].str)) {
303 if (attr_names[i].type == ATTR_NAME_TYPE_INT) {
304 char *end_ptr = NULL;
305 long long int val = strtoll(value, &end_ptr, 10);
306 if (end_ptr == value || end_ptr[0] != '\0') {
307 av_log(s, AV_LOG_ERROR,
308 "The %s field value is not a valid number: %s\n",
309 attr, value);
310 return AVERROR_INVALIDDATA;
311 }
312 if (val < attr_names[i].range.min ||
313 val > attr_names[i].range.max) {
314 av_log(s, AV_LOG_ERROR,
315 "fmtp field %s should be in range [%d,%d] (provided value: %lld)",
316 attr, attr_names[i].range.min, attr_names[i].range.max, val);
317 return AVERROR_INVALIDDATA;
318 }
319
320 *(int *)((char *)data+
321 attr_names[i].offset) = (int) val;
322 } else if (attr_names[i].type == ATTR_NAME_TYPE_STR) {
323 char *val = av_strdup(value);
324 if (!val)
325 return AVERROR(ENOMEM);
326 *(char **)((char *)data+
327 attr_names[i].offset) = val;
328 }
329 }
330 }
331 }
332 return 0;
333 }
334
parse_sdp_line(AVFormatContext * s,int st_index,PayloadContext * data,const char * line)335 static int parse_sdp_line(AVFormatContext *s, int st_index,
336 PayloadContext *data, const char *line)
337 {
338 const char *p;
339
340 if (st_index < 0)
341 return 0;
342
343 if (av_strstart(line, "fmtp:", &p))
344 return ff_parse_fmtp(s, s->streams[st_index], data, p, parse_fmtp);
345
346 return 0;
347 }
348
349 const RTPDynamicProtocolHandler ff_mp4v_es_dynamic_handler = {
350 .enc_name = "MP4V-ES",
351 .codec_type = AVMEDIA_TYPE_VIDEO,
352 .codec_id = AV_CODEC_ID_MPEG4,
353 .need_parsing = AVSTREAM_PARSE_FULL,
354 .priv_data_size = sizeof(PayloadContext),
355 .parse_sdp_a_line = parse_sdp_line,
356 };
357
358 const RTPDynamicProtocolHandler ff_mpeg4_generic_dynamic_handler = {
359 .enc_name = "mpeg4-generic",
360 .codec_type = AVMEDIA_TYPE_AUDIO,
361 .codec_id = AV_CODEC_ID_AAC,
362 .priv_data_size = sizeof(PayloadContext),
363 .parse_sdp_a_line = parse_sdp_line,
364 .close = close_context,
365 .parse_packet = aac_parse_packet,
366 };
367