gstvideodecoder.c 117 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
/* GStreamer
 * Copyright (C) 2008 David Schleef <ds@schleef.org>
 * Copyright (C) 2011 Mark Nauwelaerts <mark.nauwelaerts@collabora.co.uk>.
 * Copyright (C) 2011 Nokia Corporation. All rights reserved.
 *   Contact: Stefan Kost <stefan.kost@nokia.com>
 * Copyright (C) 2012 Collabora Ltd.
 *	Author : Edward Hervey <edward@collabora.com>
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Library General Public
 * License as published by the Free Software Foundation; either
 * version 2 of the License, or (at your option) any later version.
 *
 * This library is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Library General Public License for more details.
 *
 * You should have received a copy of the GNU Library General Public
 * License along with this library; if not, write to the
Tim-Philipp Müller's avatar
Tim-Philipp Müller committed
21
22
 * Free Software Foundation, Inc., 51 Franklin St, Fifth Floor,
 * Boston, MA 02110-1301, USA.
23
24
25
26
27
28
29
30
31
32
 */

/**
 * SECTION:gstvideodecoder
 * @short_description: Base class for video decoders
 * @see_also: 
 *
 * This base class is for video decoders turning encoded data into raw video
 * frames.
 *
33
 * The GstVideoDecoder base class and derived subclasses should cooperate as follows:
34
35
36
37
38
 * <orderedlist>
 * <listitem>
 *   <itemizedlist><title>Configuration</title>
 *   <listitem><para>
 *     Initially, GstVideoDecoder calls @start when the decoder element
39
 *     is activated, which allows the subclass to perform any global setup.
40
41
 *   </para></listitem>
 *   <listitem><para>
42
 *     GstVideoDecoder calls @set_format to inform the subclass of caps
43
44
45
46
47
48
 *     describing input video data that it is about to receive, including
 *     possibly configuration data.
 *     While unlikely, it might be called more than once, if changing input
 *     parameters require reconfiguration.
 *   </para></listitem>
 *   <listitem><para>
49
50
51
 *     Incoming data buffers are processed as needed, described in Data Processing below.
 *   </para></listitem>
 *   <listitem><para>
52
53
54
55
56
57
58
59
 *     GstVideoDecoder calls @stop at end of all processing.
 *   </para></listitem>
 *   </itemizedlist>
 * </listitem>
 * <listitem>
 *   <itemizedlist>
 *   <title>Data processing</title>
 *     <listitem><para>
60
 *       The base class gathers input data, and optionally allows subclass
61
62
63
64
 *       to parse this into subsequently manageable chunks, typically
 *       corresponding to and referred to as 'frames'.
 *     </para></listitem>
 *     <listitem><para>
65
66
 *       Each input frame is provided in turn to the subclass' @handle_frame callback.
 *       The ownership of the frame is given to the @handle_frame callback.
67
68
 *     </para></listitem>
 *     <listitem><para>
69
 *       If codec processing results in decoded data, the subclass should call
70
 *       @gst_video_decoder_finish_frame to have decoded data pushed.
71
72
73
 *       downstream. Otherwise, the subclass must call @gst_video_decoder_drop_frame, to
 *       allow the base class to do timestamp and offset tracking, and possibly to
 *       requeue the frame for a later attempt in the case of reverse playback.
74
75
76
77
78
79
 *     </para></listitem>
 *   </itemizedlist>
 * </listitem>
 * <listitem>
 *   <itemizedlist><title>Shutdown phase</title>
 *   <listitem><para>
80
 *     The GstVideoDecoder class calls @stop to inform the subclass that data
81
82
83
84
 *     parsing will be stopped.
 *   </para></listitem>
 *   </itemizedlist>
 * </listitem>
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
 * <listitem>
 *   <itemizedlist><title>Additional Notes</title>
 *   <listitem>
 *     <itemizedlist><title>Seeking/Flushing</title>
 *     <listitem><para>
 *   When the pipeline is seeked or otherwise flushed, the subclass is informed via a call
 *   to its @reset callback, with the hard parameter set to true. This indicates the
 *   subclass should drop any internal data queues and timestamps and prepare for a fresh
 *   set of buffers to arrive for parsing and decoding.
 *     </para></listitem>
 *     </itemizedlist>
 *   </listitem>
 *   <listitem>
 *     <itemizedlist><title>End Of Stream</title>
 *     <listitem><para>
 *   At end-of-stream, the subclass @parse function may be called some final times with the 
 *   at_eos parameter set to true, indicating that the element should not expect any more data
 *   to be arriving, and it should parse and remaining frames and call
 *   gst_video_decoder_have_frame() if possible.
 *     </para></listitem>
 *     </itemizedlist>
 *   </listitem>
 *   </itemizedlist>
 * </listitem>
109
110
 * </orderedlist>
 *
111
 * The subclass is responsible for providing pad template caps for
112
 * source and sink pads. The pads need to be named "sink" and "src". It also
113
114
115
116
117
 * needs to provide information about the ouptput caps, when they are known.
 * This may be when the base class calls the subclass' @set_format function,
 * though it might be during decoding, before calling
 * @gst_video_decoder_finish_frame. This is done via
 * @gst_video_decoder_set_output_state
118
 *
119
 * The subclass is also responsible for providing (presentation) timestamps
120
 * (likely based on corresponding input ones).  If that is not applicable
121
 * or possible, the base class provides limited framerate based interpolation.
122
 *
123
124
 * Similarly, the base class provides some limited (legacy) seeking support
 * if specifically requested by the subclass, as full-fledged support
125
126
 * should rather be left to upstream demuxer, parser or alike.  This simple
 * approach caters for seeking and duration reporting using estimated input
127
128
 * bitrates. To enable it, a subclass should call
 * @gst_video_decoder_set_estimate_rate to enable handling of incoming byte-streams.
129
 *
130
 * The base class provides some support for reverse playback, in particular
131
 * in case incoming data is not packetized or upstream does not provide
132
133
134
135
136
 * fragments on keyframe boundaries.  However, the subclass should then be prepared
 * for the parsing and frame processing stage to occur separately (in normal
 * forward processing, the latter immediately follows the former),
 * The subclass also needs to ensure the parsing stage properly marks keyframes,
 * unless it knows the upstream elements will do so properly for incoming data.
137
 *
138
 * The bare minimum that a functional subclass needs to implement is:
139
140
141
 * <itemizedlist>
 *   <listitem><para>Provide pad templates</para></listitem>
 *   <listitem><para>
142
 *      Inform the base class of output caps via @gst_video_decoder_set_output_state
143
144
 *   </para></listitem>
 *   <listitem><para>
145
 *      Parse input data, if it is not considered packetized from upstream
146
 *      Data will be provided to @parse which should invoke @gst_video_decoder_add_to_frame and
147
 *      @gst_video_decoder_have_frame to separate the data belonging to each video frame.
148
149
150
 *   </para></listitem>
 *   <listitem><para>
 *      Accept data in @handle_frame and provide decoded results to
151
 *      @gst_video_decoder_finish_frame, or call @gst_video_decoder_drop_frame.
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
 *   </para></listitem>
 * </itemizedlist>
 */

#ifdef HAVE_CONFIG_H
#include "config.h"
#endif

/* TODO
 *
 * * Add a flag/boolean for I-frame-only/image decoders so we can do extra
 *   features, like applying QoS on input (as opposed to after the frame is
 *   decoded).
 * * Add a flag/boolean for decoders that require keyframes, so the base
 *   class can automatically discard non-keyframes before one has arrived
167
 * * Detect reordered frame/timestamps and fix the pts/dts
168
169
170
171
172
173
174
 * * Support for GstIndex (or shall we not care ?)
 * * Calculate actual latency based on input/output timestamp/frame_number
 *   and if it exceeds the recorded one, save it and emit a GST_MESSAGE_LATENCY
 * * Emit latency message when it changes
 *
 */

175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
/* Implementation notes:
 * The Video Decoder base class operates in 2 primary processing modes, depending
 * on whether forward or reverse playback is requested.
 *
 * Forward playback:
 *   * Incoming buffer -> @parse() -> add_to_frame()/have_frame() -> handle_frame() -> 
 *     push downstream
 *
 * Reverse playback is more complicated, since it involves gathering incoming data regions
 * as we loop backwards through the upstream data. The processing concept (using incoming
 * buffers as containing one frame each to simplify things) is:
 *
 * Upstream data we want to play:
 *  Buffer encoded order:  1  2  3  4  5  6  7  8  9  EOS
 *  Keyframe flag:            K        K        
 *  Groupings:             AAAAAAA  BBBBBBB  CCCCCCC
 *
 * Input:
 *  Buffer reception order:  7  8  9  4  5  6  1  2  3  EOS
 *  Keyframe flag:                       K        K
 *  Discont flag:            D        D        D
 *
 * - Each Discont marks a discont in the decoding order.
 * - The keyframes mark where we can start decoding.
 *
 * Initially, we prepend incoming buffers to the gather queue. Whenever the
 * discont flag is set on an incoming buffer, the gather queue is flushed out
 * before the new buffer is collected.
 *
 * The above data will be accumulated in the gather queue like this:
 *
 *   gather queue:  9  8  7
 *                        D
 *
 * Whe buffer 4 is received (with a DISCONT), we flush the gather queue like
 * this:
 *
 *   while (gather)
 *     take head of queue and prepend to parse queue (this reverses the sequence,
 *     so parse queue is 7 -> 8 -> 9)
 *
 *   Next, we process the parse queue, which now contains all un-parsed packets (including
 *   any leftover ones from the previous decode section)
 *
 *   for each buffer now in the parse queue:
 *     Call the subclass parse function, prepending each resulting frame to
 *     the parse_gather queue. Buffers which precede the first one that
 *     produces a parsed frame are retained in the parse queue for re-processing on
 *     the next cycle of parsing.
 *
 *   The parse_gather queue now contains frame objects ready for decoding, in reverse order.
 *   parse_gather: 9 -> 8 -> 7
 *
 *   while (parse_gather)
 *     Take the head of the queue and prepend it to the decode queue
 *     If the frame was a keyframe, process the decode queue
 *   decode is now 7-8-9
 *
 *  Processing the decode queue results in frames with attached output buffers
 *  stored in the 'output_queue' ready for outputting in reverse order.
 *
 * After we flushed the gather queue and parsed it, we add 4 to the (now empty) gather queue.
 * We get the following situation:
 *
 *  gather queue:    4
 *  decode queue:    7  8  9
 *
 * After we received 5 (Keyframe) and 6:
 *
 *  gather queue:    6  5  4
 *  decode queue:    7  8  9
 *
 * When we receive 1 (DISCONT) which triggers a flush of the gather queue:
 *
 *   Copy head of the gather queue (6) to decode queue:
 *
 *    gather queue:    5  4
 *    decode queue:    6  7  8  9
 *
 *   Copy head of the gather queue (5) to decode queue. This is a keyframe so we
 *   can start decoding.
 *
 *    gather queue:    4
 *    decode queue:    5  6  7  8  9
 *
 *   Decode frames in decode queue, store raw decoded data in output queue, we
 *   can take the head of the decode queue and prepend the decoded result in the
 *   output queue:
 *
 *    gather queue:    4
 *    decode queue:    
 *    output queue:    9  8  7  6  5
 *
 *   Now output all the frames in the output queue, picking a frame from the
 *   head of the queue.
 *
 *   Copy head of the gather queue (4) to decode queue, we flushed the gather
 *   queue and can now store input buffer in the gather queue:
 *
 *    gather queue:    1
 *    decode queue:    4
 *
 *  When we receive EOS, the queue looks like:
 *
 *    gather queue:    3  2  1
 *    decode queue:    4
 *
 *  Fill decode queue, first keyframe we copy is 2:
 *
 *    gather queue:    1
 *    decode queue:    2  3  4
 *
 *  Decoded output:
 *
 *    gather queue:    1
 *    decode queue:    
 *    output queue:    4  3  2
 *
 *  Leftover buffer 1 cannot be decoded and must be discarded.
 */

296
297
298
#include "gstvideodecoder.h"
#include "gstvideoutils.h"

299
#include <gst/video/video.h>
300
#include <gst/video/video-event.h>
301
302
#include <gst/video/gstvideopool.h>
#include <gst/video/gstvideometa.h>
303
304
305
306
307
308
309
310
311
312
313
314
315
#include <string.h>

GST_DEBUG_CATEGORY (videodecoder_debug);
#define GST_CAT_DEFAULT videodecoder_debug

#define GST_VIDEO_DECODER_GET_PRIVATE(obj)  \
    (G_TYPE_INSTANCE_GET_PRIVATE ((obj), GST_TYPE_VIDEO_DECODER, \
        GstVideoDecoderPrivate))

struct _GstVideoDecoderPrivate
{
  /* FIXME introduce a context ? */

316
  GstBufferPool *pool;
317
318
  GstAllocator *allocator;
  GstAllocationParams params;
319

320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
  /* parse tracking */
  /* input data */
  GstAdapter *input_adapter;
  /* assembles current frame */
  GstAdapter *output_adapter;

  /* Whether we attempt to convert newsegment from bytes to
   * time using a bitrate estimation */
  gboolean do_estimate_rate;

  /* Whether input is considered packetized or not */
  gboolean packetized;

  /* Error handling */
  gint max_errors;
  gint error_count;
336
337
  gboolean had_output_data;
  gboolean had_input_data;
338

339
  gboolean needs_format;
340
341
  gboolean do_caps;

342
343
344
345
346
  /* ... being tracked here;
   * only available during parsing */
  GstVideoCodecFrame *current_frame;
  /* events that should apply to the current frame */
  GList *current_frame_events;
347
348
  /* events that should be pushed before the next frame */
  GList *pending_events;
349
350
351
352
353
354
355
356

  /* relative offset of input data */
  guint64 input_offset;
  /* relative offset of frame */
  guint64 frame_offset;
  /* tracking ts and offsets */
  GList *timestamps;

357
  /* last outgoing ts */
358
  GstClockTime last_timestamp_out;
359
360
  /* incoming pts - dts */
  GstClockTime pts_delta;
361
  gboolean reordered_output;
362
363
364
365
366
367
368
369
370
371

  /* reverse playback */
  /* collect input */
  GList *gather;
  /* to-be-parsed */
  GList *parse;
  /* collected parsed frames */
  GList *parse_gather;
  /* frames to be handled == decoded */
  GList *decode;
372
373
  /* collected output - of buffer objects, not frames */
  GList *output_queued;
374

375
376

  /* base_picture_number is the picture number of the reference picture */
377
  guint64 base_picture_number;
378
379
380
  /* combine with base_picture_number, framerate and calcs to yield (presentation) ts */
  GstClockTime base_timestamp;

381
382
383
384
  /* FIXME : reorder_depth is never set */
  int reorder_depth;
  int distance_from_sync;

385
386
  guint32 system_frame_number;
  guint32 decode_frame_number;
387
388
389

  GList *frames;                /* Protected with OBJECT_LOCK */
  GstVideoCodecState *input_state;
390
  GstVideoCodecState *output_state;     /* OBJECT_LOCK and STREAM_LOCK */
391
392
393
  gboolean output_state_changed;

  /* QoS properties */
394
395
396
  gdouble proportion;           /* OBJECT_LOCK */
  GstClockTime earliest_time;   /* OBJECT_LOCK */
  GstClockTime qos_frame_duration;      /* OBJECT_LOCK */
397
398
399
400
401
402
403
404
405
406
407
  gboolean discont;
  /* qos messages: frames dropped/processed */
  guint dropped;
  guint processed;

  /* Outgoing byte size ? */
  gint64 bytes_out;
  gint64 time;

  gint64 min_latency;
  gint64 max_latency;
408
409
410

  GstTagList *tags;
  gboolean tags_changed;
411
412
};

413
414
415
416
417
static GstElementClass *parent_class = NULL;
static void gst_video_decoder_class_init (GstVideoDecoderClass * klass);
static void gst_video_decoder_init (GstVideoDecoder * dec,
    GstVideoDecoderClass * klass);

418
419
static void gst_video_decoder_finalize (GObject * object);

420
421
422
423
424
425
426
427
428
429
430
431
432
433
static gboolean gst_video_decoder_setcaps (GstVideoDecoder * dec,
    GstCaps * caps);
static gboolean gst_video_decoder_sink_event (GstPad * pad, GstObject * parent,
    GstEvent * event);
static gboolean gst_video_decoder_src_event (GstPad * pad, GstObject * parent,
    GstEvent * event);
static GstFlowReturn gst_video_decoder_chain (GstPad * pad, GstObject * parent,
    GstBuffer * buf);
static gboolean gst_video_decoder_sink_query (GstPad * pad, GstObject * parent,
    GstQuery * query);
static GstStateChangeReturn gst_video_decoder_change_state (GstElement *
    element, GstStateChange transition);
static gboolean gst_video_decoder_src_query (GstPad * pad, GstObject * parent,
    GstQuery * query);
434
435
static void gst_video_decoder_reset (GstVideoDecoder * decoder, gboolean full,
    gboolean flush_hard);
436

437
438
static GstFlowReturn gst_video_decoder_decode_frame (GstVideoDecoder * decoder,
    GstVideoCodecFrame * frame);
439

440
441
static GstClockTime gst_video_decoder_get_frame_duration (GstVideoDecoder *
    decoder, GstVideoCodecFrame * frame);
442
443
static GstVideoCodecFrame *gst_video_decoder_new_frame (GstVideoDecoder *
    decoder);
444
445
static GstFlowReturn gst_video_decoder_clip_and_push_buf (GstVideoDecoder *
    decoder, GstBuffer * buf);
446
447
static GstFlowReturn gst_video_decoder_flush_parse (GstVideoDecoder * dec,
    gboolean at_eos);
448

449
450
static void gst_video_decoder_clear_queues (GstVideoDecoder * dec);

451
452
453
454
static gboolean gst_video_decoder_sink_event_default (GstVideoDecoder * decoder,
    GstEvent * event);
static gboolean gst_video_decoder_src_event_default (GstVideoDecoder * decoder,
    GstEvent * event);
455
456
static gboolean gst_video_decoder_decide_allocation_default (GstVideoDecoder *
    decoder, GstQuery * query);
457
458
static gboolean gst_video_decoder_propose_allocation_default (GstVideoDecoder *
    decoder, GstQuery * query);
459
static gboolean gst_video_decoder_negotiate_default (GstVideoDecoder * decoder);
460
static GstFlowReturn gst_video_decoder_parse_available (GstVideoDecoder * dec,
461
    gboolean at_eos, gboolean new_buffer);
462
463
static gboolean gst_video_decoder_negotiate_unlocked (GstVideoDecoder *
    decoder);
464
465
466
467
static gboolean gst_video_decoder_sink_query_default (GstVideoDecoder * decoder,
    GstQuery * query);
static gboolean gst_video_decoder_src_query_default (GstVideoDecoder * decoder,
    GstQuery * query);
468

469
470
471
472
/* we can't use G_DEFINE_ABSTRACT_TYPE because we need the klass in the _init
 * method to get to the padtemplates */
GType
gst_video_decoder_get_type (void)
473
{
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
  static volatile gsize type = 0;

  if (g_once_init_enter (&type)) {
    GType _type;
    static const GTypeInfo info = {
      sizeof (GstVideoDecoderClass),
      NULL,
      NULL,
      (GClassInitFunc) gst_video_decoder_class_init,
      NULL,
      NULL,
      sizeof (GstVideoDecoder),
      0,
      (GInstanceInitFunc) gst_video_decoder_init,
    };

    _type = g_type_register_static (GST_TYPE_ELEMENT,
        "GstVideoDecoder", &info, G_TYPE_FLAG_ABSTRACT);
    g_once_init_leave (&type, _type);
  }
  return type;
495
496
497
498
499
500
501
502
503
504
505
}

static void
gst_video_decoder_class_init (GstVideoDecoderClass * klass)
{
  GObjectClass *gobject_class;
  GstElementClass *gstelement_class;

  gobject_class = G_OBJECT_CLASS (klass);
  gstelement_class = GST_ELEMENT_CLASS (klass);

506
507
508
509
  GST_DEBUG_CATEGORY_INIT (videodecoder_debug, "videodecoder", 0,
      "Base Video Decoder");

  parent_class = g_type_class_peek_parent (klass);
510
511
512
513
514
515
  g_type_class_add_private (klass, sizeof (GstVideoDecoderPrivate));

  gobject_class->finalize = gst_video_decoder_finalize;

  gstelement_class->change_state =
      GST_DEBUG_FUNCPTR (gst_video_decoder_change_state);
516
517
518

  klass->sink_event = gst_video_decoder_sink_event_default;
  klass->src_event = gst_video_decoder_src_event_default;
519
  klass->decide_allocation = gst_video_decoder_decide_allocation_default;
520
  klass->propose_allocation = gst_video_decoder_propose_allocation_default;
521
  klass->negotiate = gst_video_decoder_negotiate_default;
522
523
  klass->sink_query = gst_video_decoder_sink_query_default;
  klass->src_query = gst_video_decoder_src_query_default;
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
}

static void
gst_video_decoder_init (GstVideoDecoder * decoder, GstVideoDecoderClass * klass)
{
  GstPadTemplate *pad_template;
  GstPad *pad;

  GST_DEBUG_OBJECT (decoder, "gst_video_decoder_init");

  decoder->priv = GST_VIDEO_DECODER_GET_PRIVATE (decoder);

  pad_template =
      gst_element_class_get_pad_template (GST_ELEMENT_CLASS (klass), "sink");
  g_return_if_fail (pad_template != NULL);

  decoder->sinkpad = pad = gst_pad_new_from_template (pad_template, "sink");

  gst_pad_set_chain_function (pad, GST_DEBUG_FUNCPTR (gst_video_decoder_chain));
  gst_pad_set_event_function (pad,
      GST_DEBUG_FUNCPTR (gst_video_decoder_sink_event));
  gst_pad_set_query_function (pad,
      GST_DEBUG_FUNCPTR (gst_video_decoder_sink_query));
  gst_element_add_pad (GST_ELEMENT (decoder), decoder->sinkpad);

  pad_template =
      gst_element_class_get_pad_template (GST_ELEMENT_CLASS (klass), "src");
  g_return_if_fail (pad_template != NULL);

  decoder->srcpad = pad = gst_pad_new_from_template (pad_template, "src");

  gst_pad_set_event_function (pad,
      GST_DEBUG_FUNCPTR (gst_video_decoder_src_event));
  gst_pad_set_query_function (pad,
      GST_DEBUG_FUNCPTR (gst_video_decoder_src_query));
  gst_pad_use_fixed_caps (pad);
  gst_element_add_pad (GST_ELEMENT (decoder), decoder->srcpad);

  gst_segment_init (&decoder->input_segment, GST_FORMAT_TIME);
  gst_segment_init (&decoder->output_segment, GST_FORMAT_TIME);

565
  g_rec_mutex_init (&decoder->stream_lock);
566
567
568
569

  decoder->priv->input_adapter = gst_adapter_new ();
  decoder->priv->output_adapter = gst_adapter_new ();
  decoder->priv->packetized = TRUE;
570
  decoder->priv->needs_format = FALSE;
571

572
  gst_video_decoder_reset (decoder, TRUE, TRUE);
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
}

static gboolean
gst_video_rawvideo_convert (GstVideoCodecState * state,
    GstFormat src_format, gint64 src_value,
    GstFormat * dest_format, gint64 * dest_value)
{
  gboolean res = FALSE;
  guint vidsize;
  guint fps_n, fps_d;

  g_return_val_if_fail (dest_format != NULL, FALSE);
  g_return_val_if_fail (dest_value != NULL, FALSE);

  if (src_format == *dest_format || src_value == 0 || src_value == -1) {
    *dest_value = src_value;
    return TRUE;
  }

  vidsize = GST_VIDEO_INFO_SIZE (&state->info);
  fps_n = GST_VIDEO_INFO_FPS_N (&state->info);
  fps_d = GST_VIDEO_INFO_FPS_D (&state->info);

  if (src_format == GST_FORMAT_BYTES &&
      *dest_format == GST_FORMAT_DEFAULT && vidsize) {
    /* convert bytes to frames */
    *dest_value = gst_util_uint64_scale_int (src_value, 1, vidsize);
    res = TRUE;
  } else if (src_format == GST_FORMAT_DEFAULT &&
      *dest_format == GST_FORMAT_BYTES && vidsize) {
    /* convert bytes to frames */
    *dest_value = src_value * vidsize;
    res = TRUE;
  } else if (src_format == GST_FORMAT_DEFAULT &&
      *dest_format == GST_FORMAT_TIME && fps_n) {
    /* convert frames to time */
    *dest_value = gst_util_uint64_scale (src_value, GST_SECOND * fps_d, fps_n);
    res = TRUE;
  } else if (src_format == GST_FORMAT_TIME &&
      *dest_format == GST_FORMAT_DEFAULT && fps_d) {
    /* convert time to frames */
    *dest_value = gst_util_uint64_scale (src_value, fps_n, GST_SECOND * fps_d);
    res = TRUE;
  } else if (src_format == GST_FORMAT_TIME &&
      *dest_format == GST_FORMAT_BYTES && fps_d && vidsize) {
618
    /* convert time to bytes */
619
    *dest_value = gst_util_uint64_scale (src_value,
620
        fps_n * (guint64) vidsize, GST_SECOND * fps_d);
621
622
623
    res = TRUE;
  } else if (src_format == GST_FORMAT_BYTES &&
      *dest_format == GST_FORMAT_TIME && fps_n && vidsize) {
624
    /* convert bytes to time */
625
    *dest_value = gst_util_uint64_scale (src_value,
626
        GST_SECOND * fps_d, fps_n * (guint64) vidsize);
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
    res = TRUE;
  }

  return res;
}

static gboolean
gst_video_encoded_video_convert (gint64 bytes, gint64 time,
    GstFormat src_format, gint64 src_value, GstFormat * dest_format,
    gint64 * dest_value)
{
  gboolean res = FALSE;

  g_return_val_if_fail (dest_format != NULL, FALSE);
  g_return_val_if_fail (dest_value != NULL, FALSE);

  if (G_UNLIKELY (src_format == *dest_format || src_value == 0 ||
          src_value == -1)) {
    if (dest_value)
      *dest_value = src_value;
    return TRUE;
  }

  if (bytes <= 0 || time <= 0) {
    GST_DEBUG ("not enough metadata yet to convert");
    goto exit;
  }

  switch (src_format) {
    case GST_FORMAT_BYTES:
      switch (*dest_format) {
        case GST_FORMAT_TIME:
          *dest_value = gst_util_uint64_scale (src_value, time, bytes);
          res = TRUE;
          break;
        default:
          res = FALSE;
      }
      break;
    case GST_FORMAT_TIME:
      switch (*dest_format) {
        case GST_FORMAT_BYTES:
          *dest_value = gst_util_uint64_scale (src_value, bytes, time);
          res = TRUE;
          break;
        default:
          res = FALSE;
      }
      break;
    default:
      GST_DEBUG ("unhandled conversion from %d to %d", src_format,
          *dest_format);
      res = FALSE;
  }

exit:
  return res;
}

static GstVideoCodecState *
_new_input_state (GstCaps * caps)
{
  GstVideoCodecState *state;
  GstStructure *structure;
  const GValue *codec_data;

  state = g_slice_new0 (GstVideoCodecState);
  state->ref_count = 1;
  gst_video_info_init (&state->info);
  if (G_UNLIKELY (!gst_video_info_from_caps (&state->info, caps)))
    goto parse_fail;
  state->caps = gst_caps_ref (caps);

  structure = gst_caps_get_structure (caps, 0);

  codec_data = gst_structure_get_value (structure, "codec_data");
  if (codec_data && G_VALUE_TYPE (codec_data) == GST_TYPE_BUFFER)
704
    state->codec_data = GST_BUFFER (g_value_dup_boxed (codec_data));
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734

  return state;

parse_fail:
  {
    g_slice_free (GstVideoCodecState, state);
    return NULL;
  }
}

static GstVideoCodecState *
_new_output_state (GstVideoFormat fmt, guint width, guint height,
    GstVideoCodecState * reference)
{
  GstVideoCodecState *state;

  state = g_slice_new0 (GstVideoCodecState);
  state->ref_count = 1;
  gst_video_info_init (&state->info);
  gst_video_info_set_format (&state->info, fmt, width, height);

  if (reference) {
    GstVideoInfo *tgt, *ref;

    tgt = &state->info;
    ref = &reference->info;

    /* Copy over extra fields from reference state */
    tgt->interlace_mode = ref->interlace_mode;
    tgt->flags = ref->flags;
735
736
    /* only copy values that are not unknown so that we don't override the
     * defaults. subclasses should really fill these in when they know. */
737
738
    if (ref->chroma_site)
      tgt->chroma_site = ref->chroma_site;
739
740
741
742
743
744
745
746
    if (ref->colorimetry.range)
      tgt->colorimetry.range = ref->colorimetry.range;
    if (ref->colorimetry.matrix)
      tgt->colorimetry.matrix = ref->colorimetry.matrix;
    if (ref->colorimetry.transfer)
      tgt->colorimetry.transfer = ref->colorimetry.transfer;
    if (ref->colorimetry.primaries)
      tgt->colorimetry.primaries = ref->colorimetry.primaries;
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
    GST_DEBUG ("reference par %d/%d fps %d/%d",
        ref->par_n, ref->par_d, ref->fps_n, ref->fps_d);
    tgt->par_n = ref->par_n;
    tgt->par_d = ref->par_d;
    tgt->fps_n = ref->fps_n;
    tgt->fps_d = ref->fps_d;
  }

  GST_DEBUG ("reference par %d/%d fps %d/%d",
      state->info.par_n, state->info.par_d,
      state->info.fps_n, state->info.fps_d);

  return state;
}

static gboolean
763
gst_video_decoder_setcaps (GstVideoDecoder * decoder, GstCaps * caps)
764
765
766
767
768
769
770
771
772
{
  GstVideoDecoderClass *decoder_class;
  GstVideoCodecState *state;
  gboolean ret = TRUE;

  decoder_class = GST_VIDEO_DECODER_GET_CLASS (decoder);

  GST_DEBUG_OBJECT (decoder, "setcaps %" GST_PTR_FORMAT, caps);

773
774
775
776
777
778
779
780
781
782
  GST_VIDEO_DECODER_STREAM_LOCK (decoder);

  if (decoder->priv->input_state) {
    GST_DEBUG_OBJECT (decoder,
        "Checking if caps changed old %" GST_PTR_FORMAT " new %" GST_PTR_FORMAT,
        decoder->priv->input_state->caps, caps);
    if (gst_caps_is_equal (decoder->priv->input_state->caps, caps))
      goto caps_not_changed;
  }

783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
  state = _new_input_state (caps);

  if (G_UNLIKELY (state == NULL))
    goto parse_fail;

  if (decoder_class->set_format)
    ret = decoder_class->set_format (decoder, state);

  if (!ret)
    goto refused_format;

  if (decoder->priv->input_state)
    gst_video_codec_state_unref (decoder->priv->input_state);
  decoder->priv->input_state = state;

  GST_VIDEO_DECODER_STREAM_UNLOCK (decoder);

  return ret;

802
803
804
805
806
807
caps_not_changed:
  {
    GST_DEBUG_OBJECT (decoder, "Caps did not change - ignore");
    GST_VIDEO_DECODER_STREAM_UNLOCK (decoder);
    return TRUE;
  }
808

809
  /* ERRORS */
810
811
812
parse_fail:
  {
    GST_WARNING_OBJECT (decoder, "Failed to parse caps");
813
    GST_VIDEO_DECODER_STREAM_UNLOCK (decoder);
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
    return FALSE;
  }

refused_format:
  {
    GST_VIDEO_DECODER_STREAM_UNLOCK (decoder);
    GST_WARNING_OBJECT (decoder, "Subclass refused caps");
    gst_video_codec_state_unref (state);
    return FALSE;
  }
}

static void
gst_video_decoder_finalize (GObject * object)
{
  GstVideoDecoder *decoder;

  decoder = GST_VIDEO_DECODER (object);

  GST_DEBUG_OBJECT (object, "finalize");

835
  g_rec_mutex_clear (&decoder->stream_lock);
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850

  if (decoder->priv->input_adapter) {
    g_object_unref (decoder->priv->input_adapter);
    decoder->priv->input_adapter = NULL;
  }
  if (decoder->priv->output_adapter) {
    g_object_unref (decoder->priv->output_adapter);
    decoder->priv->output_adapter = NULL;
  }

  if (decoder->priv->input_state)
    gst_video_codec_state_unref (decoder->priv->input_state);
  if (decoder->priv->output_state)
    gst_video_codec_state_unref (decoder->priv->output_state);

851
  if (decoder->priv->pool) {
852
853
854
855
856
    gst_object_unref (decoder->priv->pool);
    decoder->priv->pool = NULL;
  }

  if (decoder->priv->allocator) {
Wim Taymans's avatar
Wim Taymans committed
857
    gst_object_unref (decoder->priv->allocator);
858
    decoder->priv->allocator = NULL;
859
860
  }

861
862
863
864
865
866
867
  G_OBJECT_CLASS (parent_class)->finalize (object);
}

/* hard == FLUSH, otherwise discont */
static GstFlowReturn
gst_video_decoder_flush (GstVideoDecoder * dec, gboolean hard)
{
868
  GstVideoDecoderClass *klass = GST_VIDEO_DECODER_GET_CLASS (dec);
869
870
871
872
  GstFlowReturn ret = GST_FLOW_OK;

  GST_LOG_OBJECT (dec, "flush hard %d", hard);

873
874
875
876
877
878
  /* Inform subclass */
  if (klass->reset) {
    GST_FIXME_OBJECT (dec, "GstVideoDecoder::reset() is deprecated");
    klass->reset (dec, hard);
  }

879
  if (klass->flush)
880
    klass->flush (dec);
881

882
  /* and get (re)set for the sequel */
883
  gst_video_decoder_reset (dec, FALSE, hard);
884
885
886
887

  return ret;
}

888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
static gboolean
gst_video_decoder_push_event (GstVideoDecoder * decoder, GstEvent * event)
{
  switch (GST_EVENT_TYPE (event)) {
    case GST_EVENT_SEGMENT:
    {
      GstSegment segment;

      GST_VIDEO_DECODER_STREAM_LOCK (decoder);

      gst_event_copy_segment (event, &segment);

      GST_DEBUG_OBJECT (decoder, "segment %" GST_SEGMENT_FORMAT, &segment);

      if (segment.format != GST_FORMAT_TIME) {
        GST_DEBUG_OBJECT (decoder, "received non TIME newsegment");
        GST_VIDEO_DECODER_STREAM_UNLOCK (decoder);
        break;
      }

      decoder->output_segment = segment;
      GST_VIDEO_DECODER_STREAM_UNLOCK (decoder);
      break;
    }
    default:
      break;
  }

916
917
918
  GST_DEBUG_OBJECT (decoder, "pushing event %s",
      gst_event_type_get_name (GST_EVENT_TYPE (event)));

919
920
921
  return gst_pad_push_event (decoder->srcpad, event);
}

922
static GstFlowReturn
923
924
gst_video_decoder_parse_available (GstVideoDecoder * dec, gboolean at_eos,
    gboolean new_buffer)
925
926
927
928
929
930
931
932
933
{
  GstVideoDecoderClass *decoder_class = GST_VIDEO_DECODER_GET_CLASS (dec);
  GstVideoDecoderPrivate *priv = dec->priv;
  GstFlowReturn ret = GST_FLOW_OK;
  gsize start_size, available;

  available = gst_adapter_available (priv->input_adapter);
  start_size = 0;

934
935
936
  while (ret == GST_FLOW_OK && ((available && start_size != available)
          || new_buffer)) {
    new_buffer = FALSE;
937
938
939
940
941
942
943
944
945
946
947
948
949
950
    /* current frame may have been parsed and handled,
     * so we need to set up a new one when asking subclass to parse */
    if (priv->current_frame == NULL)
      priv->current_frame = gst_video_decoder_new_frame (dec);

    start_size = available;
    ret = decoder_class->parse (dec, priv->current_frame,
        priv->input_adapter, at_eos);
    available = gst_adapter_available (priv->input_adapter);
  }

  return ret;
}

951
static GstFlowReturn
952
gst_video_decoder_drain_out (GstVideoDecoder * dec, gboolean at_eos)
953
954
955
956
957
958
959
960
961
962
963
{
  GstVideoDecoderClass *decoder_class = GST_VIDEO_DECODER_GET_CLASS (dec);
  GstVideoDecoderPrivate *priv = dec->priv;
  GstFlowReturn ret = GST_FLOW_OK;

  GST_VIDEO_DECODER_STREAM_LOCK (dec);

  if (dec->input_segment.rate > 0.0) {
    /* Forward mode, if unpacketized, give the child class
     * a final chance to flush out packets */
    if (!priv->packetized) {
964
      ret = gst_video_decoder_parse_available (dec, TRUE, FALSE);
965
    }
966
967
968
969
970

    if (at_eos) {
      if (decoder_class->finish)
        ret = decoder_class->finish (dec);
    }
971
972
973
974
975
976
977
978
979
980
  } else {
    /* Reverse playback mode */
    ret = gst_video_decoder_flush_parse (dec, TRUE);
  }

  GST_VIDEO_DECODER_STREAM_UNLOCK (dec);

  return ret;
}

981
static gboolean
982
983
gst_video_decoder_sink_event_default (GstVideoDecoder * decoder,
    GstEvent * event)
984
985
{
  GstVideoDecoderPrivate *priv;
986
  gboolean ret = FALSE;
987
  gboolean forward_immediate = FALSE;
988
989
990
991

  priv = decoder->priv;

  switch (GST_EVENT_TYPE (event)) {
992
993
994
995
996
997
998
    case GST_EVENT_STREAM_START:
    {
      GstFlowReturn flow_ret = GST_FLOW_OK;

      flow_ret = gst_video_decoder_drain_out (decoder, FALSE);
      ret = (flow_ret == GST_FLOW_OK);

999
1000
      GST_DEBUG_OBJECT (decoder, "received STREAM_START. Clearing taglist");
      GST_VIDEO_DECODER_STREAM_LOCK (decoder);