1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-11-26 19:01:44 +02:00
FFmpeg/fftools/ffmpeg.c

2969 lines
97 KiB
C
Raw Normal View History

/*
* Copyright (c) 2000-2003 Fabrice Bellard
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg 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
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file
* multimedia converter based on the FFmpeg libraries
*/
#include "config.h"
#include <errno.h>
#include <limits.h>
#include <stdatomic.h>
#include <stdint.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#if HAVE_IO_H
#include <io.h>
#endif
#if HAVE_UNISTD_H
#include <unistd.h>
#endif
#if HAVE_SYS_RESOURCE_H
#include <sys/time.h>
#include <sys/types.h>
#include <sys/resource.h>
#elif HAVE_GETPROCESSTIMES
#include <windows.h>
#endif
#if HAVE_GETPROCESSMEMORYINFO
#include <windows.h>
#include <psapi.h>
#endif
#if HAVE_SETCONSOLECTRLHANDLER
#include <windows.h>
#endif
#if HAVE_SYS_SELECT_H
#include <sys/select.h>
#endif
#if HAVE_TERMIOS_H
#include <fcntl.h>
#include <sys/ioctl.h>
#include <sys/time.h>
#include <termios.h>
#elif HAVE_KBHIT
#include <conio.h>
#endif
#include "libavutil/avassert.h"
#include "libavutil/avstring.h"
#include "libavutil/bprint.h"
#include "libavutil/channel_layout.h"
#include "libavutil/dict.h"
#include "libavutil/display.h"
#include "libavutil/fifo.h"
#include "libavutil/hwcontext.h"
#include "libavutil/imgutils.h"
#include "libavutil/intreadwrite.h"
#include "libavutil/libm.h"
#include "libavutil/mathematics.h"
#include "libavutil/opt.h"
#include "libavutil/parseutils.h"
#include "libavutil/pixdesc.h"
#include "libavutil/samplefmt.h"
#include "libavutil/thread.h"
#include "libavutil/threadmessage.h"
#include "libavutil/time.h"
#include "libavutil/timestamp.h"
#include "libavcodec/version.h"
#include "libavformat/avformat.h"
#include "libavdevice/avdevice.h"
#include "libswresample/swresample.h"
#include "libavfilter/avfilter.h"
#include "libavfilter/buffersrc.h"
#include "libavfilter/buffersink.h"
#include "cmdutils.h"
#include "ffmpeg.h"
fftools/ffmpeg: rework -shortest implementation The -shortest option (which finishes the output file at the time the shortest stream ends) is currently implemented by faking the -t option when an output stream ends. This approach is fragile, since it depends on the frames/packets being processed in a specific order. E.g. there are currently some situations in which the output file length will depend unpredictably on unrelated factors like encoder delay. More importantly, the present work aiming at splitting various ffmpeg components into different threads will make this approach completely unworkable, since the frames/packets will arrive in effectively random order. This commit introduces a "sync queue", which is essentially a collection of FIFOs, one per stream. Frames/packets are submitted to these FIFOs and are then released for further processing (encoding or muxing) when it is ensured that the frame in question will not cause its stream to get ahead of the other streams (the logic is similar to libavformat's interleaving queue). These sync queues are then used for encoding and/or muxing when the -shortest option is specified. A new option – -shortest_buf_duration – controls the maximum number of queued packets, to avoid runaway memory usage. This commit changes the results of the following tests: - copy-shortest[12]: the last audio frame is now gone. This is correct, since it actually outlasts the last video frame. - shortest-sub: the video packets following the last subtitle packet are now gone. This is also correct.
2022-06-10 14:21:42 +02:00
#include "sync_queue.h"
2011-04-23 16:19:17 +03:00
const char program_name[] = "ffmpeg";
const int program_birth_year = 2000;
FILE *vstats_file;
typedef struct BenchmarkTimeStamps {
int64_t real_usec;
int64_t user_usec;
int64_t sys_usec;
} BenchmarkTimeStamps;
static BenchmarkTimeStamps get_benchmark_time_stamps(void);
static int64_t getmaxrss(void);
int64_t nb_frames_dup = 0;
int64_t nb_frames_drop = 0;
static int64_t decode_error_stat[2];
unsigned nb_output_dumped = 0;
static BenchmarkTimeStamps current_time;
AVIOContext *progress_avio = NULL;
InputFile **input_files = NULL;
int nb_input_files = 0;
OutputFile **output_files = NULL;
int nb_output_files = 0;
FilterGraph **filtergraphs;
int nb_filtergraphs;
#if HAVE_TERMIOS_H
/* init terminal so that we can grab keys */
static struct termios oldtty;
static int restore_tty;
#endif
/* sub2video hack:
Convert subtitles to video with alpha to insert them in filter graphs.
This is a temporary solution until libavfilter gets real subtitles support.
*/
static int sub2video_get_blank_frame(InputStream *ist)
{
int ret;
AVFrame *frame = ist->sub2video.frame;
av_frame_unref(frame);
ist->sub2video.frame->width = ist->dec_ctx->width ? ist->dec_ctx->width : ist->sub2video.w;
ist->sub2video.frame->height = ist->dec_ctx->height ? ist->dec_ctx->height : ist->sub2video.h;
ist->sub2video.frame->format = AV_PIX_FMT_RGB32;
if ((ret = av_frame_get_buffer(frame, 0)) < 0)
return ret;
memset(frame->data[0], 0, frame->height * frame->linesize[0]);
return 0;
}
static void sub2video_copy_rect(uint8_t *dst, int dst_linesize, int w, int h,
AVSubtitleRect *r)
{
uint32_t *pal, *dst2;
uint8_t *src, *src2;
int x, y;
if (r->type != SUBTITLE_BITMAP) {
av_log(NULL, AV_LOG_WARNING, "sub2video: non-bitmap subtitle\n");
return;
}
if (r->x < 0 || r->x + r->w > w || r->y < 0 || r->y + r->h > h) {
av_log(NULL, AV_LOG_WARNING, "sub2video: rectangle (%d %d %d %d) overflowing %d %d\n",
r->x, r->y, r->w, r->h, w, h
);
return;
}
dst += r->y * dst_linesize + r->x * 4;
src = r->data[0];
pal = (uint32_t *)r->data[1];
for (y = 0; y < r->h; y++) {
dst2 = (uint32_t *)dst;
src2 = src;
for (x = 0; x < r->w; x++)
*(dst2++) = pal[*(src2++)];
dst += dst_linesize;
src += r->linesize[0];
}
}
static void sub2video_push_ref(InputStream *ist, int64_t pts)
{
AVFrame *frame = ist->sub2video.frame;
int i;
int ret;
av_assert1(frame->data[0]);
ist->sub2video.last_pts = frame->pts = pts;
for (i = 0; i < ist->nb_filters; i++) {
ret = av_buffersrc_add_frame_flags(ist->filters[i]->filter, frame,
AV_BUFFERSRC_FLAG_KEEP_REF |
AV_BUFFERSRC_FLAG_PUSH);
if (ret != AVERROR_EOF && ret < 0)
av_log(NULL, AV_LOG_WARNING, "Error while add the frame to buffer source(%s).\n",
av_err2str(ret));
}
}
void sub2video_update(InputStream *ist, int64_t heartbeat_pts, AVSubtitle *sub)
{
AVFrame *frame = ist->sub2video.frame;
int8_t *dst;
int dst_linesize;
int num_rects, i;
int64_t pts, end_pts;
if (!frame)
return;
if (sub) {
pts = av_rescale_q(sub->pts + sub->start_display_time * 1000LL,
AV_TIME_BASE_Q, ist->st->time_base);
end_pts = av_rescale_q(sub->pts + sub->end_display_time * 1000LL,
AV_TIME_BASE_Q, ist->st->time_base);
num_rects = sub->num_rects;
} else {
/* If we are initializing the system, utilize current heartbeat
PTS as the start time, and show until the following subpicture
is received. Otherwise, utilize the previous subpicture's end time
as the fall-back value. */
pts = ist->sub2video.initialize ?
heartbeat_pts : ist->sub2video.end_pts;
end_pts = INT64_MAX;
num_rects = 0;
}
if (sub2video_get_blank_frame(ist) < 0) {
av_log(NULL, AV_LOG_ERROR,
"Impossible to get a blank canvas.\n");
return;
}
dst = frame->data [0];
dst_linesize = frame->linesize[0];
for (i = 0; i < num_rects; i++)
sub2video_copy_rect(dst, dst_linesize, frame->width, frame->height, sub->rects[i]);
sub2video_push_ref(ist, pts);
ist->sub2video.end_pts = end_pts;
ist->sub2video.initialize = 0;
}
static void sub2video_heartbeat(InputStream *ist, int64_t pts)
{
InputFile *infile = input_files[ist->file_index];
int i, j, nb_reqs;
int64_t pts2;
/* When a frame is read from a file, examine all sub2video streams in
the same file and send the sub2video frame again. Otherwise, decoded
video frames could be accumulating in the filter graph while a filter
(possibly overlay) is desperately waiting for a subtitle frame. */
for (i = 0; i < infile->nb_streams; i++) {
InputStream *ist2 = infile->streams[i];
if (!ist2->sub2video.frame)
continue;
/* subtitles seem to be usually muxed ahead of other streams;
if not, subtracting a larger time here is necessary */
pts2 = av_rescale_q(pts, ist->st->time_base, ist2->st->time_base) - 1;
/* do not send the heartbeat frame if the subtitle is already ahead */
if (pts2 <= ist2->sub2video.last_pts)
continue;
if (pts2 >= ist2->sub2video.end_pts || ist2->sub2video.initialize)
/* if we have hit the end of the current displayed subpicture,
or if we need to initialize the system, update the
overlayed subpicture and its start/end times */
sub2video_update(ist2, pts2 + 1, NULL);
for (j = 0, nb_reqs = 0; j < ist2->nb_filters; j++)
nb_reqs += av_buffersrc_get_nb_failed_requests(ist2->filters[j]->filter);
if (nb_reqs)
sub2video_push_ref(ist2, pts2);
}
}
static void sub2video_flush(InputStream *ist)
{
int i;
int ret;
if (ist->sub2video.end_pts < INT64_MAX)
sub2video_update(ist, INT64_MAX, NULL);
for (i = 0; i < ist->nb_filters; i++) {
ret = av_buffersrc_add_frame(ist->filters[i]->filter, NULL);
if (ret != AVERROR_EOF && ret < 0)
av_log(NULL, AV_LOG_WARNING, "Flush the frame error.\n");
}
}
/* end of sub2video hack */
static void term_exit_sigsafe(void)
{
#if HAVE_TERMIOS_H
if(restore_tty)
tcsetattr (0, TCSANOW, &oldtty);
#endif
}
void term_exit(void)
{
av_log(NULL, AV_LOG_QUIET, "%s", "");
term_exit_sigsafe();
}
static volatile int received_sigterm = 0;
static volatile int received_nb_signals = 0;
static atomic_int transcode_init_done = ATOMIC_VAR_INIT(0);
static volatile int ffmpeg_exited = 0;
static int64_t copy_ts_first_pts = AV_NOPTS_VALUE;
static void
sigterm_handler(int sig)
{
int ret;
received_sigterm = sig;
received_nb_signals++;
term_exit_sigsafe();
if(received_nb_signals > 3) {
ret = write(2/*STDERR_FILENO*/, "Received > 3 system signals, hard exiting\n",
strlen("Received > 3 system signals, hard exiting\n"));
if (ret < 0) { /* Do nothing */ };
exit(123);
}
}
#if HAVE_SETCONSOLECTRLHANDLER
static BOOL WINAPI CtrlHandler(DWORD fdwCtrlType)
{
av_log(NULL, AV_LOG_DEBUG, "\nReceived windows signal %ld\n", fdwCtrlType);
switch (fdwCtrlType)
{
case CTRL_C_EVENT:
case CTRL_BREAK_EVENT:
sigterm_handler(SIGINT);
return TRUE;
case CTRL_CLOSE_EVENT:
case CTRL_LOGOFF_EVENT:
case CTRL_SHUTDOWN_EVENT:
sigterm_handler(SIGTERM);
/* Basically, with these 3 events, when we return from this method the
process is hard terminated, so stall as long as we need to
to try and let the main thread(s) clean up and gracefully terminate
(we have at most 5 seconds, but should be done far before that). */
while (!ffmpeg_exited) {
Sleep(0);
}
return TRUE;
default:
av_log(NULL, AV_LOG_ERROR, "Received unknown windows signal %ld\n", fdwCtrlType);
return FALSE;
}
}
#endif
#ifdef __linux__
#define SIGNAL(sig, func) \
do { \
action.sa_handler = func; \
sigaction(sig, &action, NULL); \
} while (0)
#else
#define SIGNAL(sig, func) \
signal(sig, func)
#endif
void term_init(void)
{
#if defined __linux__
struct sigaction action = {0};
action.sa_handler = sigterm_handler;
/* block other interrupts while processing this one */
sigfillset(&action.sa_mask);
/* restart interruptible functions (i.e. don't fail with EINTR) */
action.sa_flags = SA_RESTART;
#endif
#if HAVE_TERMIOS_H
if (stdin_interaction) {
struct termios tty;
if (tcgetattr (0, &tty) == 0) {
oldtty = tty;
restore_tty = 1;
tty.c_iflag &= ~(IGNBRK|BRKINT|PARMRK|ISTRIP
|INLCR|IGNCR|ICRNL|IXON);
tty.c_oflag |= OPOST;
tty.c_lflag &= ~(ECHO|ECHONL|ICANON|IEXTEN);
tty.c_cflag &= ~(CSIZE|PARENB);
tty.c_cflag |= CS8;
tty.c_cc[VMIN] = 1;
tty.c_cc[VTIME] = 0;
tcsetattr (0, TCSANOW, &tty);
}
SIGNAL(SIGQUIT, sigterm_handler); /* Quit (POSIX). */
}
#endif
SIGNAL(SIGINT , sigterm_handler); /* Interrupt (ANSI). */
SIGNAL(SIGTERM, sigterm_handler); /* Termination (ANSI). */
#ifdef SIGXCPU
SIGNAL(SIGXCPU, sigterm_handler);
#endif
#ifdef SIGPIPE
signal(SIGPIPE, SIG_IGN); /* Broken pipe (POSIX). */
#endif
#if HAVE_SETCONSOLECTRLHANDLER
SetConsoleCtrlHandler((PHANDLER_ROUTINE) CtrlHandler, TRUE);
#endif
}
/* read a key without blocking */
static int read_key(void)
{
unsigned char ch;
#if HAVE_TERMIOS_H
int n = 1;
struct timeval tv;
fd_set rfds;
FD_ZERO(&rfds);
FD_SET(0, &rfds);
tv.tv_sec = 0;
tv.tv_usec = 0;
n = select(1, &rfds, NULL, NULL, &tv);
if (n > 0) {
n = read(0, &ch, 1);
if (n == 1)
return ch;
return n;
}
#elif HAVE_KBHIT
# if HAVE_PEEKNAMEDPIPE
static int is_pipe;
static HANDLE input_handle;
DWORD dw, nchars;
if(!input_handle){
input_handle = GetStdHandle(STD_INPUT_HANDLE);
is_pipe = !GetConsoleMode(input_handle, &dw);
}
Merge remote-tracking branch 'qatar/master' * qatar/master: (27 commits) libxvid: Give more suitable names to libxvid-related files. libxvid: Separate libxvid encoder from libxvid rate control code. jpeglsdec: Remove write-only variable in ff_jpegls_decode_lse(). fate: cosmetics: lowercase some comments fate: Give more consistent names to some RealVideo/RealAudio tests. lavfi: add avfilter_get_audio_buffer_ref_from_arrays(). lavfi: add extended_data to AVFilterBuffer. lavc: check that extended_data is properly set in avcodec_encode_audio2(). lavc: pad last audio frame with silence when needed. samplefmt: add a function for filling a buffer with silence. samplefmt: add a function for copying audio samples. lavr: do not try to copy to uninitialized output audio data. lavr: make avresample_read() with NULL output discard samples. fate: split idroq audio and video into separate tests fate: improve dependencies fate: add convenient shorthands for ea-vp6, libavcodec, libavutil tests fate: split some combined tests into separate audio and video tests fate: fix dependencies for probe tests mips: intreadwrite: fix inline asm for gcc 4.8 mips: intreadwrite: remove unnecessary inline asm ... Conflicts: cmdutils.h configure doc/APIchanges doc/filters.texi ffmpeg.c ffplay.c libavcodec/internal.h libavcodec/jpeglsdec.c libavcodec/libschroedingerdec.c libavcodec/libxvid.c libavcodec/libxvid_rc.c libavcodec/utils.c libavcodec/version.h libavfilter/avfilter.c libavfilter/avfilter.h libavfilter/buffersink.h tests/Makefile tests/fate/aac.mak tests/fate/audio.mak tests/fate/demux.mak tests/fate/ea.mak tests/fate/image.mak tests/fate/libavutil.mak tests/fate/lossless-audio.mak tests/fate/lossless-video.mak tests/fate/microsoft.mak tests/fate/qt.mak tests/fate/real.mak tests/fate/screen.mak tests/fate/video.mak tests/fate/voice.mak tests/fate/vqf.mak tests/ref/fate/ea-mad tests/ref/fate/ea-tqi Merged-by: Michael Niedermayer <michaelni@gmx.at>
2012-05-10 03:14:44 +03:00
if (is_pipe) {
/* When running under a GUI, you will end here. */
if (!PeekNamedPipe(input_handle, NULL, 0, NULL, &nchars, NULL)) {
// input pipe may have been closed by the program that ran ffmpeg
return -1;
}
//Read it
if(nchars != 0) {
read(0, &ch, 1);
return ch;
}else{
return -1;
}
}
# endif
if(kbhit())
return(getch());
#endif
return -1;
}
static int decode_interrupt_cb(void *ctx)
{
return received_nb_signals > atomic_load(&transcode_init_done);
}
const AVIOInterruptCB int_cb = { decode_interrupt_cb, NULL };
static void ffmpeg_cleanup(int ret)
{
int i, j;
if (do_benchmark) {
int maxrss = getmaxrss() / 1024;
av_log(NULL, AV_LOG_INFO, "bench: maxrss=%ikB\n", maxrss);
}
for (i = 0; i < nb_filtergraphs; i++) {
FilterGraph *fg = filtergraphs[i];
avfilter_graph_free(&fg->graph);
for (j = 0; j < fg->nb_inputs; j++) {
InputFilter *ifilter = fg->inputs[j];
struct InputStream *ist = ifilter->ist;
if (ifilter->frame_queue) {
AVFrame *frame;
while (av_fifo_read(ifilter->frame_queue, &frame, 1) >= 0)
av_frame_free(&frame);
av_fifo_freep2(&ifilter->frame_queue);
}
av_freep(&ifilter->displaymatrix);
if (ist->sub2video.sub_queue) {
2022-01-10 12:04:36 +02:00
AVSubtitle sub;
while (av_fifo_read(ist->sub2video.sub_queue, &sub, 1) >= 0)
avsubtitle_free(&sub);
2022-01-10 12:04:36 +02:00
av_fifo_freep2(&ist->sub2video.sub_queue);
}
av_buffer_unref(&ifilter->hw_frames_ctx);
av_freep(&ifilter->name);
av_freep(&fg->inputs[j]);
}
av_freep(&fg->inputs);
for (j = 0; j < fg->nb_outputs; j++) {
OutputFilter *ofilter = fg->outputs[j];
avfilter_inout_free(&ofilter->out_tmp);
av_freep(&ofilter->name);
av_channel_layout_uninit(&ofilter->ch_layout);
av_freep(&fg->outputs[j]);
}
av_freep(&fg->outputs);
av_freep(&fg->graph_desc);
av_freep(&filtergraphs[i]);
}
av_freep(&filtergraphs);
/* close files */
for (i = 0; i < nb_output_files; i++)
of_close(&output_files[i]);
for (i = 0; i < nb_input_files; i++)
ifile_close(&input_files[i]);
if (vstats_file) {
if (fclose(vstats_file))
av_log(NULL, AV_LOG_ERROR,
"Error closing vstats file, loss of information possible: %s\n",
av_err2str(AVERROR(errno)));
}
av_freep(&vstats_filename);
of_enc_stats_close();
hw_device_free_all();
av_freep(&filter_nbthreads);
av_freep(&input_files);
av_freep(&output_files);
uninit_opts();
avformat_network_deinit();
if (received_sigterm) {
av_log(NULL, AV_LOG_INFO, "Exiting normally, received signal %d.\n",
(int) received_sigterm);
} else if (ret && atomic_load(&transcode_init_done)) {
av_log(NULL, AV_LOG_INFO, "Conversion failed!\n");
}
term_exit();
ffmpeg_exited = 1;
}
OutputStream *ost_iter(OutputStream *prev)
{
int of_idx = prev ? prev->file_index : 0;
int ost_idx = prev ? prev->index + 1 : 0;
for (; of_idx < nb_output_files; of_idx++) {
OutputFile *of = output_files[of_idx];
if (ost_idx < of->nb_streams)
return of->streams[ost_idx];
ost_idx = 0;
}
return NULL;
}
InputStream *ist_iter(InputStream *prev)
{
int if_idx = prev ? prev->file_index : 0;
int ist_idx = prev ? prev->st->index + 1 : 0;
for (; if_idx < nb_input_files; if_idx++) {
InputFile *f = input_files[if_idx];
if (ist_idx < f->nb_streams)
return f->streams[ist_idx];
ist_idx = 0;
}
return NULL;
}
void remove_avoptions(AVDictionary **a, AVDictionary *b)
{
const AVDictionaryEntry *t = NULL;
while ((t = av_dict_iterate(b, t))) {
av_dict_set(a, t->key, NULL, AV_DICT_MATCH_CASE);
}
}
void assert_avoptions(AVDictionary *m)
{
const AVDictionaryEntry *t;
if ((t = av_dict_get(m, "", NULL, AV_DICT_IGNORE_SUFFIX))) {
av_log(NULL, AV_LOG_FATAL, "Option %s not found.\n", t->key);
exit_program(1);
}
}
static void abort_codec_experimental(const AVCodec *c, int encoder)
{
exit_program(1);
}
void update_benchmark(const char *fmt, ...)
{
if (do_benchmark_all) {
BenchmarkTimeStamps t = get_benchmark_time_stamps();
va_list va;
char buf[1024];
if (fmt) {
va_start(va, fmt);
vsnprintf(buf, sizeof(buf), fmt, va);
va_end(va);
av_log(NULL, AV_LOG_INFO,
"bench: %8" PRIu64 " user %8" PRIu64 " sys %8" PRIu64 " real %s \n",
t.user_usec - current_time.user_usec,
t.sys_usec - current_time.sys_usec,
t.real_usec - current_time.real_usec, buf);
}
current_time = t;
}
}
void close_output_stream(OutputStream *ost)
{
OutputFile *of = output_files[ost->file_index];
ost->finished |= ENCODER_FINISHED;
fftools/ffmpeg: rework -shortest implementation The -shortest option (which finishes the output file at the time the shortest stream ends) is currently implemented by faking the -t option when an output stream ends. This approach is fragile, since it depends on the frames/packets being processed in a specific order. E.g. there are currently some situations in which the output file length will depend unpredictably on unrelated factors like encoder delay. More importantly, the present work aiming at splitting various ffmpeg components into different threads will make this approach completely unworkable, since the frames/packets will arrive in effectively random order. This commit introduces a "sync queue", which is essentially a collection of FIFOs, one per stream. Frames/packets are submitted to these FIFOs and are then released for further processing (encoding or muxing) when it is ensured that the frame in question will not cause its stream to get ahead of the other streams (the logic is similar to libavformat's interleaving queue). These sync queues are then used for encoding and/or muxing when the -shortest option is specified. A new option – -shortest_buf_duration – controls the maximum number of queued packets, to avoid runaway memory usage. This commit changes the results of the following tests: - copy-shortest[12]: the last audio frame is now gone. This is correct, since it actually outlasts the last video frame. - shortest-sub: the video packets following the last subtitle packet are now gone. This is also correct.
2022-06-10 14:21:42 +02:00
if (ost->sq_idx_encode >= 0)
sq_send(of->sq_encode, ost->sq_idx_encode, SQFRAME(NULL));
}
/**
* Get and encode new output from any of the filtergraphs, without causing
* activity.
*
* @return 0 for success, <0 for severe errors
*/
static int reap_filters(int flush)
{
AVFrame *filtered_frame = NULL;
2012-08-19 12:23:59 +03:00
/* Reap all buffers present in the buffer sinks */
for (OutputStream *ost = ost_iter(NULL); ost; ost = ost_iter(ost)) {
AVFilterContext *filter;
AVCodecContext *enc = ost->enc_ctx;
2012-08-19 12:23:59 +03:00
int ret = 0;
if (!ost->filter || !ost->filter->graph->graph)
2012-08-19 12:23:59 +03:00
continue;
filter = ost->filter->filter;
2012-08-19 12:23:59 +03:00
filtered_frame = ost->filtered_frame;
while (1) {
ret = av_buffersink_get_frame_flags(filter, filtered_frame,
2012-08-19 12:23:59 +03:00
AV_BUFFERSINK_FLAG_NO_REQUEST);
if (ret < 0) {
if (ret != AVERROR(EAGAIN) && ret != AVERROR_EOF) {
av_log(NULL, AV_LOG_WARNING,
"Error in av_buffersink_get_frame_flags(): %s\n", av_err2str(ret));
} else if (flush && ret == AVERROR_EOF) {
2016-12-18 13:34:35 +02:00
if (av_buffersink_get_type(filter) == AVMEDIA_TYPE_VIDEO)
enc_frame(ost, NULL);
}
2012-08-19 12:23:59 +03:00
break;
}
if (ost->finished) {
av_frame_unref(filtered_frame);
continue;
}
if (filtered_frame->pts != AV_NOPTS_VALUE) {
AVRational tb = av_buffersink_get_time_base(filter);
ost->filter->last_pts = av_rescale_q(filtered_frame->pts, tb,
AV_TIME_BASE_Q);
filtered_frame->time_base = tb;
if (debug_ts)
av_log(NULL, AV_LOG_INFO, "filter_raw -> pts:%s pts_time:%s time_base:%d/%d\n",
av_ts2str(filtered_frame->pts),
av_ts2timestr(filtered_frame->pts, &tb),
tb.num, tb.den);
}
2016-12-18 13:34:35 +02:00
switch (av_buffersink_get_type(filter)) {
2012-08-19 12:23:59 +03:00
case AVMEDIA_TYPE_VIDEO:
if (!ost->frame_aspect_ratio.num)
enc->sample_aspect_ratio = filtered_frame->sample_aspect_ratio;
2012-08-19 12:23:59 +03:00
enc_frame(ost, filtered_frame);
2012-08-19 12:23:59 +03:00
break;
case AVMEDIA_TYPE_AUDIO:
if (!(enc->codec->capabilities & AV_CODEC_CAP_PARAM_CHANGE) &&
avcodec_is_open(enc) &&
enc->ch_layout.nb_channels != filtered_frame->ch_layout.nb_channels) {
av_log(NULL, AV_LOG_ERROR,
"Audio filter graph output is not normalized and encoder does not support parameter changes\n");
break;
}
enc_frame(ost, filtered_frame);
2012-08-19 12:23:59 +03:00
break;
default:
// TODO support subtitle filters
av_assert0(0);
}
av_frame_unref(filtered_frame);
}
2012-08-19 12:23:59 +03:00
}
return 0;
}
static void print_report(int is_last_report, int64_t timer_start, int64_t cur_time)
{
AVBPrint buf, buf_script;
int64_t total_size = of_filesize(output_files[0]);
int vid;
double bitrate;
double speed;
int64_t pts = INT64_MIN + 1;
static int64_t last_time = -1;
static int first_report = 1;
int hours, mins, secs, us;
const char *hours_sign;
int ret;
float t;
if (!print_stats && !is_last_report && !progress_avio)
return;
if (!is_last_report) {
if (last_time == -1) {
last_time = cur_time;
}
if (((cur_time - last_time) < stats_period && !first_report) ||
(first_report && nb_output_dumped < nb_output_files))
return;
last_time = cur_time;
}
t = (cur_time-timer_start) / 1000000.0;
vid = 0;
av_bprint_init(&buf, 0, AV_BPRINT_SIZE_AUTOMATIC);
av_bprint_init(&buf_script, 0, AV_BPRINT_SIZE_AUTOMATIC);
for (OutputStream *ost = ost_iter(NULL); ost; ost = ost_iter(ost)) {
const AVCodecContext * const enc = ost->enc_ctx;
const float q = enc ? ost->quality / (float) FF_QP2LAMBDA : -1;
if (vid && ost->st->codecpar->codec_type == AVMEDIA_TYPE_VIDEO) {
av_bprintf(&buf, "q=%2.1f ", q);
av_bprintf(&buf_script, "stream_%d_%d_q=%.1f\n",
ost->file_index, ost->index, q);
}
if (!vid && ost->st->codecpar->codec_type == AVMEDIA_TYPE_VIDEO) {
float fps;
uint64_t frame_number = atomic_load(&ost->packets_written);
fps = t > 1 ? frame_number / t : 0;
av_bprintf(&buf, "frame=%5"PRId64" fps=%3.*f q=%3.1f ",
frame_number, fps < 9.95, fps, q);
av_bprintf(&buf_script, "frame=%"PRId64"\n", frame_number);
av_bprintf(&buf_script, "fps=%.2f\n", fps);
av_bprintf(&buf_script, "stream_%d_%d_q=%.1f\n",
ost->file_index, ost->index, q);
if (is_last_report)
av_bprintf(&buf, "L");
if (enc && (enc->flags & AV_CODEC_FLAG_PSNR) &&
(ost->pict_type != AV_PICTURE_TYPE_NONE || is_last_report)) {
int j;
double error, error_sum = 0;
double scale, scale_sum = 0;
double p;
char type[3] = { 'Y','U','V' };
av_bprintf(&buf, "PSNR=");
for (j = 0; j < 3; j++) {
if (is_last_report) {
error = enc->error[j];
scale = enc->width * enc->height * 255.0 * 255.0 * frame_number;
} else {
error = ost->error[j];
scale = enc->width * enc->height * 255.0 * 255.0;
}
if (j)
scale /= 4;
error_sum += error;
scale_sum += scale;
p = psnr(error / scale);
av_bprintf(&buf, "%c:%2.2f ", type[j], p);
av_bprintf(&buf_script, "stream_%d_%d_psnr_%c=%2.2f\n",
ost->file_index, ost->index, type[j] | 32, p);
}
p = psnr(error_sum / scale_sum);
av_bprintf(&buf, "*:%2.2f ", psnr(error_sum / scale_sum));
av_bprintf(&buf_script, "stream_%d_%d_psnr_all=%2.2f\n",
ost->file_index, ost->index, p);
}
vid = 1;
}
/* compute min output value */
if (ost->last_mux_dts != AV_NOPTS_VALUE) {
pts = FFMAX(pts, ost->last_mux_dts);
if (copy_ts) {
if (copy_ts_first_pts == AV_NOPTS_VALUE && pts > 1)
copy_ts_first_pts = pts;
if (copy_ts_first_pts != AV_NOPTS_VALUE)
pts -= copy_ts_first_pts;
}
}
if (is_last_report)
nb_frames_drop += ost->last_dropped;
}
secs = FFABS(pts) / AV_TIME_BASE;
us = FFABS(pts) % AV_TIME_BASE;
mins = secs / 60;
secs %= 60;
hours = mins / 60;
mins %= 60;
hours_sign = (pts < 0) ? "-" : "";
bitrate = pts && total_size >= 0 ? total_size * 8 / (pts / 1000.0) : -1;
speed = t != 0.0 ? (double)pts / AV_TIME_BASE / t : -1;
if (total_size < 0) av_bprintf(&buf, "size=N/A time=");
else av_bprintf(&buf, "size=%8.0fkB time=", total_size / 1024.0);
if (pts == AV_NOPTS_VALUE) {
av_bprintf(&buf, "N/A ");
} else {
av_bprintf(&buf, "%s%02d:%02d:%02d.%02d ",
hours_sign, hours, mins, secs, (100 * us) / AV_TIME_BASE);
}
if (bitrate < 0) {
av_bprintf(&buf, "bitrate=N/A");
av_bprintf(&buf_script, "bitrate=N/A\n");
}else{
av_bprintf(&buf, "bitrate=%6.1fkbits/s", bitrate);
av_bprintf(&buf_script, "bitrate=%6.1fkbits/s\n", bitrate);
}
if (total_size < 0) av_bprintf(&buf_script, "total_size=N/A\n");
else av_bprintf(&buf_script, "total_size=%"PRId64"\n", total_size);
if (pts == AV_NOPTS_VALUE) {
av_bprintf(&buf_script, "out_time_us=N/A\n");
av_bprintf(&buf_script, "out_time_ms=N/A\n");
av_bprintf(&buf_script, "out_time=N/A\n");
} else {
av_bprintf(&buf_script, "out_time_us=%"PRId64"\n", pts);
av_bprintf(&buf_script, "out_time_ms=%"PRId64"\n", pts);
av_bprintf(&buf_script, "out_time=%s%02d:%02d:%02d.%06d\n",
hours_sign, hours, mins, secs, us);
}
if (nb_frames_dup || nb_frames_drop)
av_bprintf(&buf, " dup=%"PRId64" drop=%"PRId64, nb_frames_dup, nb_frames_drop);
av_bprintf(&buf_script, "dup_frames=%"PRId64"\n", nb_frames_dup);
av_bprintf(&buf_script, "drop_frames=%"PRId64"\n", nb_frames_drop);
if (speed < 0) {
av_bprintf(&buf, " speed=N/A");
av_bprintf(&buf_script, "speed=N/A\n");
} else {
av_bprintf(&buf, " speed=%4.3gx", speed);
av_bprintf(&buf_script, "speed=%4.3gx\n", speed);
}
if (print_stats || is_last_report) {
const char end = is_last_report ? '\n' : '\r';
if (print_stats==1 && AV_LOG_INFO > av_log_get_level()) {
fprintf(stderr, "%s %c", buf.str, end);
} else
av_log(NULL, AV_LOG_INFO, "%s %c", buf.str, end);
fflush(stderr);
}
av_bprint_finalize(&buf, NULL);
if (progress_avio) {
av_bprintf(&buf_script, "progress=%s\n",
is_last_report ? "end" : "continue");
avio_write(progress_avio, buf_script.str,
FFMIN(buf_script.len, buf_script.size - 1));
avio_flush(progress_avio);
av_bprint_finalize(&buf_script, NULL);
if (is_last_report) {
if ((ret = avio_closep(&progress_avio)) < 0)
av_log(NULL, AV_LOG_ERROR,
"Error closing progress log, loss of information possible: %s\n", av_err2str(ret));
}
}
first_report = 0;
}
int ifilter_parameters_from_codecpar(InputFilter *ifilter, AVCodecParameters *par)
{
int ret;
// We never got any input. Set a fake format, which will
// come from libavformat.
ifilter->format = par->format;
ifilter->sample_rate = par->sample_rate;
ifilter->width = par->width;
ifilter->height = par->height;
ifilter->sample_aspect_ratio = par->sample_aspect_ratio;
ret = av_channel_layout_copy(&ifilter->ch_layout, &par->ch_layout);
if (ret < 0)
return ret;
return 0;
}
/**
* @param dts predicted packet dts in AV_TIME_BASE_Q
*/
static void do_streamcopy(InputStream *ist, OutputStream *ost,
const AVPacket *pkt, int64_t dts)
{
OutputFile *of = output_files[ost->file_index];
int64_t start_time = (of->start_time == AV_NOPTS_VALUE) ? 0 : of->start_time;
int64_t ost_tb_start_time = av_rescale_q(start_time, AV_TIME_BASE_Q, ost->mux_timebase);
AVPacket *opkt = ost->pkt;
av_packet_unref(opkt);
// EOF: flush output bitstream filters.
if (!pkt) {
of_output_packet(of, opkt, ost, 1);
return;
}
if (!ost->streamcopy_started && !(pkt->flags & AV_PKT_FLAG_KEY) &&
!ost->copy_initial_nonkeyframes)
return;
if (!ost->streamcopy_started) {
if (!ost->copy_prior_start &&
(pkt->pts == AV_NOPTS_VALUE ?
dts < ost->ts_copy_start :
pkt->pts < av_rescale_q(ost->ts_copy_start, AV_TIME_BASE_Q, pkt->time_base)))
return;
if (of->start_time != AV_NOPTS_VALUE && dts < of->start_time)
return;
}
if (of->recording_time != INT64_MAX &&
dts >= of->recording_time + start_time) {
close_output_stream(ost);
return;
}
if (av_packet_ref(opkt, pkt) < 0)
fftools/ffmpeg: Improve streamcopy do_streamcopy() has a packet that gets zero-initialized first, then gets initialized via av_init_packet() after which some of its fields are oerwritten again with the actually desired values (unless it's EOF): The side data is copied into the packet with av_copy_packet_side_data() and if the source packet is refcounted, the packet will get a new reference to the source packet's data. Furthermore, the flags are copied and the timestamp related fields are overwritten with new values. This commit replaces this by using av_packet_ref() to both initialize the packet as well as populate its fields with the right values (unless it's EOF again in which case the packet will still be initialized). The differences to the current approach are as follows: a) There is no call to a deprecated function (av_copy_packet_side_data()) any more. b) Several fields that weren't copied before are now copied from the source packet to the new packet (e.g. pos). Some of them (the timestamp related fields) may be immediately overwritten again and some don't seem to be used at all (e.g. pos), but in return using av_packet_ref() allows to forgo the initializations. c) There was no check for whether copying side data fails or not. This has been changed: Now the program is exited in this case. Using av_packet_ref() does not lead to unnecessary copying of data, because the source packets are already always refcounted (they originate from av_read_frame()). Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2019-10-11 06:06:58 +02:00
exit_program(1);
opkt->time_base = ost->mux_timebase;
if (pkt->pts != AV_NOPTS_VALUE)
opkt->pts = av_rescale_q(pkt->pts, pkt->time_base, opkt->time_base) - ost_tb_start_time;
if (pkt->dts == AV_NOPTS_VALUE) {
opkt->dts = av_rescale_q(dts, AV_TIME_BASE_Q, opkt->time_base);
} else if (ost->st->codecpar->codec_type == AVMEDIA_TYPE_AUDIO) {
int duration = av_get_audio_frame_duration2(ist->par, pkt->size);
if(!duration)
duration = ist->par->frame_size;
opkt->dts = av_rescale_delta(pkt->time_base, pkt->dts,
(AVRational){1, ist->par->sample_rate}, duration,
&ist->filter_in_rescale_delta_last, opkt->time_base);
/* dts will be set immediately afterwards to what pts is now */
opkt->pts = opkt->dts - ost_tb_start_time;
} else
opkt->dts = av_rescale_q(pkt->dts, pkt->time_base, opkt->time_base);
opkt->dts -= ost_tb_start_time;
opkt->duration = av_rescale_q(pkt->duration, pkt->time_base, opkt->time_base);
{
int ret = trigger_fix_sub_duration_heartbeat(ost, pkt);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR,
"Subtitle heartbeat logic failed in %s! (%s)\n",
__func__, av_err2str(ret));
exit_program(1);
}
}
of_output_packet(of, opkt, ost, 0);
ost->streamcopy_started = 1;
}
static void check_decode_result(InputStream *ist, int *got_output, int ret)
{
if (*got_output || ret<0)
decode_error_stat[ret<0] ++;
if (ret < 0 && exit_on_error)
exit_program(1);
if (*got_output && ist) {
if (ist->decoded_frame->decode_error_flags || (ist->decoded_frame->flags & AV_FRAME_FLAG_CORRUPT)) {
av_log(NULL, exit_on_error ? AV_LOG_FATAL : AV_LOG_WARNING,
"%s: corrupt decoded frame in stream %d\n", input_files[ist->file_index]->ctx->url, ist->st->index);
if (exit_on_error)
exit_program(1);
}
}
}
// Filters can be configured only if the formats of all inputs are known.
int ifilter_has_all_input_formats(FilterGraph *fg)
{
int i;
for (i = 0; i < fg->nb_inputs; i++) {
if (fg->inputs[i]->format < 0 && (fg->inputs[i]->type == AVMEDIA_TYPE_AUDIO ||
fg->inputs[i]->type == AVMEDIA_TYPE_VIDEO))
return 0;
}
return 1;
}
static int ifilter_send_frame(InputFilter *ifilter, AVFrame *frame, int keep_reference)
{
FilterGraph *fg = ifilter->graph;
AVFrameSideData *sd;
int need_reinit, ret;
int buffersrc_flags = AV_BUFFERSRC_FLAG_PUSH;
if (keep_reference)
buffersrc_flags |= AV_BUFFERSRC_FLAG_KEEP_REF;
/* determine if the parameters for this input changed */
need_reinit = ifilter->format != frame->format;
switch (ifilter->ist->par->codec_type) {
case AVMEDIA_TYPE_AUDIO:
need_reinit |= ifilter->sample_rate != frame->sample_rate ||
av_channel_layout_compare(&ifilter->ch_layout, &frame->ch_layout);
break;
case AVMEDIA_TYPE_VIDEO:
need_reinit |= ifilter->width != frame->width ||
ifilter->height != frame->height;
break;
}
if (!ifilter->ist->reinit_filters && fg->graph)
need_reinit = 0;
if (!!ifilter->hw_frames_ctx != !!frame->hw_frames_ctx ||
(ifilter->hw_frames_ctx && ifilter->hw_frames_ctx->data != frame->hw_frames_ctx->data))
need_reinit = 1;
if (sd = av_frame_get_side_data(frame, AV_FRAME_DATA_DISPLAYMATRIX)) {
if (!ifilter->displaymatrix || memcmp(sd->data, ifilter->displaymatrix, sizeof(int32_t) * 9))
need_reinit = 1;
} else if (ifilter->displaymatrix)
need_reinit = 1;
if (need_reinit) {
ret = ifilter_parameters_from_frame(ifilter, frame);
if (ret < 0)
return ret;
}
/* (re)init the graph if possible, otherwise buffer the frame and return */
if (need_reinit || !fg->graph) {
if (!ifilter_has_all_input_formats(fg)) {
AVFrame *tmp = av_frame_clone(frame);
if (!tmp)
return AVERROR(ENOMEM);
2022-01-10 12:04:36 +02:00
ret = av_fifo_write(ifilter->frame_queue, &tmp, 1);
if (ret < 0)
av_frame_free(&tmp);
return ret;
}
ret = reap_filters(1);
if (ret < 0 && ret != AVERROR_EOF) {
av_log(NULL, AV_LOG_ERROR, "Error while filtering: %s\n", av_err2str(ret));
return ret;
}
ret = configure_filtergraph(fg);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Error reinitializing filters!\n");
return ret;
}
}
ret = av_buffersrc_add_frame_flags(ifilter->filter, frame, buffersrc_flags);
if (ret < 0) {
if (ret != AVERROR_EOF)
av_log(NULL, AV_LOG_ERROR, "Error while filtering: %s\n", av_err2str(ret));
return ret;
}
return 0;
}
2017-04-06 10:40:12 +02:00
static int ifilter_send_eof(InputFilter *ifilter, int64_t pts)
{
int ret;
ifilter->eof = 1;
if (ifilter->filter) {
2017-04-06 10:40:12 +02:00
ret = av_buffersrc_close(ifilter->filter, pts, AV_BUFFERSRC_FLAG_PUSH);
if (ret < 0)
return ret;
} else {
// the filtergraph was never configured
if (ifilter->format < 0) {
ret = ifilter_parameters_from_codecpar(ifilter, ifilter->ist->par);
if (ret < 0)
return ret;
}
if (ifilter->format < 0 && (ifilter->type == AVMEDIA_TYPE_AUDIO || ifilter->type == AVMEDIA_TYPE_VIDEO)) {
av_log(NULL, AV_LOG_ERROR, "Cannot determine format of input stream %d:%d after EOF\n", ifilter->ist->file_index, ifilter->ist->st->index);
return AVERROR_INVALIDDATA;
}
}
return 0;
}
// This does not quite work like avcodec_decode_audio4/avcodec_decode_video2.
// There is the following difference: if you got a frame, you must call
// it again with pkt=NULL. pkt==NULL is treated differently from pkt->size==0
// (pkt==NULL means get more output, pkt->size==0 is a flush/drain packet)
static int decode(InputStream *ist, AVCodecContext *avctx,
AVFrame *frame, int *got_frame, AVPacket *pkt)
{
int ret;
*got_frame = 0;
if (pkt) {
ret = avcodec_send_packet(avctx, pkt);
// In particular, we don't expect AVERROR(EAGAIN), because we read all
// decoded frames with avcodec_receive_frame() until done.
if (ret < 0 && ret != AVERROR_EOF)
return ret;
}
ret = avcodec_receive_frame(avctx, frame);
if (ret < 0 && ret != AVERROR(EAGAIN))
return ret;
if (ret >= 0) {
if (ist->want_frame_data) {
FrameData *fd;
av_assert0(!frame->opaque_ref);
frame->opaque_ref = av_buffer_allocz(sizeof(*fd));
if (!frame->opaque_ref) {
av_frame_unref(frame);
return AVERROR(ENOMEM);
}
fd = (FrameData*)frame->opaque_ref->data;
fd->pts = frame->pts;
fd->tb = avctx->pkt_timebase;
fd->idx = avctx->frame_num - 1;
}
*got_frame = 1;
}
return 0;
}
static int send_frame_to_filters(InputStream *ist, AVFrame *decoded_frame)
{
int i, ret;
av_assert1(ist->nb_filters > 0); /* ensure ret is initialized */
for (i = 0; i < ist->nb_filters; i++) {
ret = ifilter_send_frame(ist->filters[i], decoded_frame, i < ist->nb_filters - 1);
if (ret == AVERROR_EOF)
ret = 0; /* ignore */
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR,
"Failed to inject frame into filter network: %s\n", av_err2str(ret));
break;
}
}
return ret;
}
static int decode_audio(InputStream *ist, AVPacket *pkt, int *got_output,
int *decode_failed)
{
AVFrame *decoded_frame = ist->decoded_frame;
AVCodecContext *avctx = ist->dec_ctx;
int ret, err = 0;
AVRational decoded_frame_tb;
update_benchmark(NULL);
ret = decode(ist, avctx, decoded_frame, got_output, pkt);
update_benchmark("decode_audio %d.%d", ist->file_index, ist->st->index);
if (ret < 0)
*decode_failed = 1;
if (ret != AVERROR_EOF)
check_decode_result(ist, got_output, ret);
if (!*got_output || ret < 0)
return ret;
ist->samples_decoded += decoded_frame->nb_samples;
ist->frames_decoded++;
/* increment next_dts to use for the case where the input stream does not
have timestamps or there are multiple frames in the packet */
ist->next_pts += ((int64_t)AV_TIME_BASE * decoded_frame->nb_samples) /
decoded_frame->sample_rate;
ist->next_dts += ((int64_t)AV_TIME_BASE * decoded_frame->nb_samples) /
decoded_frame->sample_rate;
if (decoded_frame->pts != AV_NOPTS_VALUE) {
decoded_frame_tb = ist->st->time_base;
} else if (pkt && pkt->pts != AV_NOPTS_VALUE) {
decoded_frame->pts = pkt->pts;
decoded_frame_tb = ist->st->time_base;
}else {
decoded_frame->pts = ist->dts;
decoded_frame_tb = AV_TIME_BASE_Q;
}
if (pkt && pkt->duration && ist->prev_pkt_pts != AV_NOPTS_VALUE &&
pkt->pts != AV_NOPTS_VALUE && pkt->pts - ist->prev_pkt_pts > pkt->duration)
ist->filter_in_rescale_delta_last = AV_NOPTS_VALUE;
if (pkt)
ist->prev_pkt_pts = pkt->pts;
if (decoded_frame->pts != AV_NOPTS_VALUE)
decoded_frame->pts = av_rescale_delta(decoded_frame_tb, decoded_frame->pts,
(AVRational){1, decoded_frame->sample_rate},
decoded_frame->nb_samples,
&ist->filter_in_rescale_delta_last,
(AVRational){1, decoded_frame->sample_rate});
ist->nb_samples = decoded_frame->nb_samples;
err = send_frame_to_filters(ist, decoded_frame);
av_frame_unref(decoded_frame);
return err < 0 ? err : ret;
}
static int64_t video_duration_estimate(const InputStream *ist, const AVFrame *frame)
{
const InputFile *ifile = input_files[ist->file_index];
const int container_nots = !!(ifile->ctx->iformat->flags & AVFMT_NOTIMESTAMPS);
int64_t codec_duration = 0;
// XXX lavf currently makes up frame durations when they are not provided by
// the container. As there is no way to reliably distinguish real container
// durations from the fake made-up ones, we use heuristics based on whether
// the container has timestamps. Eventually lavf should stop making up
// durations, then this should be simplified.
// prefer frame duration for containers with timestamps
if (frame->duration > 0 && !container_nots)
return frame->duration;
if (ist->dec_ctx->framerate.den && ist->dec_ctx->framerate.num) {
int ticks = frame->repeat_pict >= 0 ?
frame->repeat_pict + 1 :
ist->dec_ctx->ticks_per_frame;
codec_duration = av_rescale_q(ticks, av_inv_q(ist->dec_ctx->framerate),
ist->st->time_base);
}
// prefer codec-layer duration for containers without timestamps
if (codec_duration > 0 && container_nots)
return codec_duration;
// when timestamps are available, repeat last frame's actual duration
// (i.e. pts difference between this and last frame)
if (frame->pts != AV_NOPTS_VALUE && ist->last_frame_pts != AV_NOPTS_VALUE &&
frame->pts > ist->last_frame_pts)
return frame->pts - ist->last_frame_pts;
// try frame/codec duration
if (frame->duration > 0)
return frame->duration;
if (codec_duration > 0)
return codec_duration;
// try average framerate
if (ist->st->avg_frame_rate.num && ist->st->avg_frame_rate.den) {
int64_t d = av_rescale_q(1, av_inv_q(ist->st->avg_frame_rate),
ist->st->time_base);
if (d > 0)
return d;
}
// last resort is last frame's estimated duration, and 1
return FFMAX(ist->last_frame_duration_est, 1);
}
static int decode_video(InputStream *ist, AVPacket *pkt, int *got_output, int64_t *duration_pts, int eof,
int *decode_failed)
{
AVFrame *decoded_frame = ist->decoded_frame;
int ret = 0, err = 0;
int64_t best_effort_timestamp;
// With fate-indeo3-2, we're getting 0-sized packets before EOF for some
// reason. This seems like a semi-critical bug. Don't trigger EOF, and
// skip the packet.
if (!eof && pkt && pkt->size == 0)
return 0;
update_benchmark(NULL);
ret = decode(ist, ist->dec_ctx, decoded_frame, got_output, pkt);
update_benchmark("decode_video %d.%d", ist->file_index, ist->st->index);
if (ret < 0)
*decode_failed = 1;
// The following line may be required in some cases where there is no parser
// or the parser does not has_b_frames correctly
if (ist->par->video_delay < ist->dec_ctx->has_b_frames) {
if (ist->dec_ctx->codec_id == AV_CODEC_ID_H264) {
ist->par->video_delay = ist->dec_ctx->has_b_frames;
} else
av_log(ist->dec_ctx, AV_LOG_WARNING,
"video_delay is larger in decoder than demuxer %d > %d.\n"
"If you want to help, upload a sample "
"of this file to https://streams.videolan.org/upload/ "
"and contact the ffmpeg-devel mailing list. (ffmpeg-devel@ffmpeg.org)\n",
ist->dec_ctx->has_b_frames,
ist->par->video_delay);
}
if (ret != AVERROR_EOF)
check_decode_result(ist, got_output, ret);
if (*got_output && ret >= 0) {
if (ist->dec_ctx->width != decoded_frame->width ||
ist->dec_ctx->height != decoded_frame->height ||
ist->dec_ctx->pix_fmt != decoded_frame->format) {
av_log(NULL, AV_LOG_DEBUG, "Frame parameters mismatch context %d,%d,%d != %d,%d,%d\n",
decoded_frame->width,
decoded_frame->height,
decoded_frame->format,
ist->dec_ctx->width,
ist->dec_ctx->height,
ist->dec_ctx->pix_fmt);
}
}
if (!*got_output || ret < 0)
return ret;
if(ist->top_field_first>=0)
decoded_frame->top_field_first = ist->top_field_first;
ist->frames_decoded++;
if (ist->hwaccel_retrieve_data && decoded_frame->format == ist->hwaccel_pix_fmt) {
err = ist->hwaccel_retrieve_data(ist->dec_ctx, decoded_frame);
if (err < 0)
goto fail;
}
best_effort_timestamp= decoded_frame->best_effort_timestamp;
*duration_pts = decoded_frame->duration;
if (ist->framerate.num)
best_effort_timestamp = ist->cfr_next_pts++;
// no timestamp available - extrapolate from previous frame duration
if (best_effort_timestamp == AV_NOPTS_VALUE &&
ist->last_frame_pts != AV_NOPTS_VALUE)
best_effort_timestamp = ist->last_frame_pts + ist->last_frame_duration_est;
if (best_effort_timestamp == AV_NOPTS_VALUE)
best_effort_timestamp = av_rescale_q(ist->pts, AV_TIME_BASE_Q, ist->st->time_base);
if(best_effort_timestamp != AV_NOPTS_VALUE) {
int64_t ts = av_rescale_q(decoded_frame->pts = best_effort_timestamp, ist->st->time_base, AV_TIME_BASE_Q);
if (ts != AV_NOPTS_VALUE)
ist->next_pts = ist->pts = ts;
}
// update timestamp history
ist->last_frame_duration_est = video_duration_estimate(ist, decoded_frame);
ist->last_frame_pts = decoded_frame->pts;
if (debug_ts) {
av_log(ist, AV_LOG_INFO,
"decoder -> pts:%s pts_time:%s "
"pkt_dts:%s pkt_dts_time:%s "
"best_effort_ts:%"PRId64" best_effort_ts_time:%s "
"duration:%s duration_time:%s "
"keyframe:%d frame_type:%d time_base:%d/%d\n",
av_ts2str(decoded_frame->pts),
av_ts2timestr(decoded_frame->pts, &ist->st->time_base),
av_ts2str(decoded_frame->pkt_dts),
av_ts2timestr(decoded_frame->pkt_dts, &ist->st->time_base),
best_effort_timestamp,
av_ts2timestr(best_effort_timestamp, &ist->st->time_base),
av_ts2str(decoded_frame->duration),
av_ts2timestr(decoded_frame->duration, &ist->st->time_base),
decoded_frame->key_frame, decoded_frame->pict_type,
ist->st->time_base.num, ist->st->time_base.den);
}
if (ist->st->sample_aspect_ratio.num)
decoded_frame->sample_aspect_ratio = ist->st->sample_aspect_ratio;
err = send_frame_to_filters(ist, decoded_frame);
fail:
av_frame_unref(decoded_frame);
return err < 0 ? err : ret;
}
static int process_subtitle(InputStream *ist, AVSubtitle *subtitle, int *got_output)
{
int ret = 0;
int free_sub = 1;
if (ist->fix_sub_duration) {
int end = 1;
if (ist->prev_sub.got_output) {
end = av_rescale(subtitle->pts - ist->prev_sub.subtitle.pts,
1000, AV_TIME_BASE);
if (end < ist->prev_sub.subtitle.end_display_time) {
av_log(NULL, AV_LOG_DEBUG,
"Subtitle duration reduced from %"PRId32" to %d%s\n",
ist->prev_sub.subtitle.end_display_time, end,
end <= 0 ? ", dropping it" : "");
ist->prev_sub.subtitle.end_display_time = end;
}
}
FFSWAP(int, *got_output, ist->prev_sub.got_output);
FFSWAP(int, ret, ist->prev_sub.ret);
FFSWAP(AVSubtitle, *subtitle, ist->prev_sub.subtitle);
if (end <= 0)
goto out;
}
if (!*got_output)
return ret;
if (ist->sub2video.frame) {
sub2video_update(ist, INT64_MIN, subtitle);
} else if (ist->nb_filters) {
if (!ist->sub2video.sub_queue)
2022-01-10 12:04:36 +02:00
ist->sub2video.sub_queue = av_fifo_alloc2(8, sizeof(AVSubtitle), AV_FIFO_FLAG_AUTO_GROW);
if (!ist->sub2video.sub_queue)
report_and_exit(AVERROR(ENOMEM));
2022-01-10 12:04:36 +02:00
ret = av_fifo_write(ist->sub2video.sub_queue, subtitle, 1);
2022-01-10 12:04:36 +02:00
if (ret < 0)
exit_program(1);
free_sub = 0;
}
if (!subtitle->num_rects)
goto out;
for (int oidx = 0; oidx < ist->nb_outputs; oidx++) {
OutputStream *ost = ist->outputs[oidx];
if (!ost->enc_ctx || ost->enc_ctx->codec_type != AVMEDIA_TYPE_SUBTITLE)
continue;
enc_subtitle(output_files[ost->file_index], ost, subtitle);
}
out:
if (free_sub)
avsubtitle_free(subtitle);
return ret;
}
static int copy_av_subtitle(AVSubtitle *dst, AVSubtitle *src)
{
int ret = AVERROR_BUG;
AVSubtitle tmp = {
.format = src->format,
.start_display_time = src->start_display_time,
.end_display_time = src->end_display_time,
.num_rects = 0,
.rects = NULL,
.pts = src->pts
};
if (!src->num_rects)
goto success;
if (!(tmp.rects = av_calloc(src->num_rects, sizeof(*tmp.rects))))
return AVERROR(ENOMEM);
for (int i = 0; i < src->num_rects; i++) {
AVSubtitleRect *src_rect = src->rects[i];
AVSubtitleRect *dst_rect;
if (!(dst_rect = tmp.rects[i] = av_mallocz(sizeof(*tmp.rects[0])))) {
ret = AVERROR(ENOMEM);
goto cleanup;
}
tmp.num_rects++;
dst_rect->type = src_rect->type;
dst_rect->flags = src_rect->flags;
dst_rect->x = src_rect->x;
dst_rect->y = src_rect->y;
dst_rect->w = src_rect->w;
dst_rect->h = src_rect->h;
dst_rect->nb_colors = src_rect->nb_colors;
if (src_rect->text)
if (!(dst_rect->text = av_strdup(src_rect->text))) {
ret = AVERROR(ENOMEM);
goto cleanup;
}
if (src_rect->ass)
if (!(dst_rect->ass = av_strdup(src_rect->ass))) {
ret = AVERROR(ENOMEM);
goto cleanup;
}
for (int j = 0; j < 4; j++) {
// SUBTITLE_BITMAP images are special in the sense that they
// are like PAL8 images. first pointer to data, second to
// palette. This makes the size calculation match this.
size_t buf_size = src_rect->type == SUBTITLE_BITMAP && j == 1 ?
AVPALETTE_SIZE :
src_rect->h * src_rect->linesize[j];
if (!src_rect->data[j])
continue;
if (!(dst_rect->data[j] = av_memdup(src_rect->data[j], buf_size))) {
ret = AVERROR(ENOMEM);
goto cleanup;
}
dst_rect->linesize[j] = src_rect->linesize[j];
}
}
success:
*dst = tmp;
return 0;
cleanup:
avsubtitle_free(&tmp);
return ret;
}
static int fix_sub_duration_heartbeat(InputStream *ist, int64_t signal_pts)
{
int ret = AVERROR_BUG;
int got_output = 1;
AVSubtitle *prev_subtitle = &ist->prev_sub.subtitle;
AVSubtitle subtitle;
if (!ist->fix_sub_duration || !prev_subtitle->num_rects ||
signal_pts <= prev_subtitle->pts)
return 0;
if ((ret = copy_av_subtitle(&subtitle, prev_subtitle)) < 0)
return ret;
subtitle.pts = signal_pts;
return process_subtitle(ist, &subtitle, &got_output);
}
int trigger_fix_sub_duration_heartbeat(OutputStream *ost, const AVPacket *pkt)
{
OutputFile *of = output_files[ost->file_index];
int64_t signal_pts = av_rescale_q(pkt->pts, pkt->time_base,
AV_TIME_BASE_Q);
if (!ost->fix_sub_duration_heartbeat || !(pkt->flags & AV_PKT_FLAG_KEY))
// we are only interested in heartbeats on streams configured, and
// only on random access points.
return 0;
for (int i = 0; i < of->nb_streams; i++) {
OutputStream *iter_ost = of->streams[i];
InputStream *ist = iter_ost->ist;
int ret = AVERROR_BUG;
if (iter_ost == ost || !ist || !ist->decoding_needed ||
ist->dec_ctx->codec_type != AVMEDIA_TYPE_SUBTITLE)
// We wish to skip the stream that causes the heartbeat,
// output streams without an input stream, streams not decoded
// (as fix_sub_duration is only done for decoded subtitles) as
// well as non-subtitle streams.
continue;
if ((ret = fix_sub_duration_heartbeat(ist, signal_pts)) < 0)
return ret;
}
return 0;
}
static int transcode_subtitles(InputStream *ist, const AVPacket *pkt,
int *got_output, int *decode_failed)
{
AVSubtitle subtitle;
int ret = avcodec_decode_subtitle2(ist->dec_ctx,
&subtitle, got_output, pkt);
check_decode_result(NULL, got_output, ret);
if (ret < 0 || !*got_output) {
*decode_failed = 1;
if (!pkt->size)
sub2video_flush(ist);
return ret;
}
ist->frames_decoded++;
return process_subtitle(ist, &subtitle, got_output);
}
static int send_filter_eof(InputStream *ist)
{
int i, ret;
2017-04-06 10:40:12 +02:00
/* TODO keep pts also in stream time base to avoid converting back */
int64_t pts = av_rescale_q_rnd(ist->pts, AV_TIME_BASE_Q, ist->st->time_base,
AV_ROUND_NEAR_INF | AV_ROUND_PASS_MINMAX);
for (i = 0; i < ist->nb_filters; i++) {
2017-04-06 10:40:12 +02:00
ret = ifilter_send_eof(ist->filters[i], pts);
if (ret < 0)
return ret;
}
return 0;
}
/* pkt = NULL means EOF (needed to flush decoder buffers) */
static int process_input_packet(InputStream *ist, const AVPacket *pkt, int no_eof)
{
InputFile *f = input_files[ist->file_index];
const AVCodecParameters *par = ist->par;
int ret = 0;
int repeating = 0;
int eof_reached = 0;
int duration_exceeded;
AVPacket *avpkt = ist->pkt;
if (!ist->saw_first_ts) {
ist->first_dts =
ist->dts = ist->st->avg_frame_rate.num ? - ist->dec_ctx->has_b_frames * AV_TIME_BASE / av_q2d(ist->st->avg_frame_rate) : 0;
ist->pts = 0;
if (pkt && pkt->pts != AV_NOPTS_VALUE && !ist->decoding_needed) {
ist->first_dts =
ist->dts += av_rescale_q(pkt->pts, ist->st->time_base, AV_TIME_BASE_Q);
}
ist->saw_first_ts = 1;
}
if (ist->next_dts == AV_NOPTS_VALUE)
ist->next_dts = ist->dts;
if (ist->next_pts == AV_NOPTS_VALUE)
ist->next_pts = ist->pts;
if (pkt) {
av_packet_unref(avpkt);
ret = av_packet_ref(avpkt, pkt);
if (ret < 0)
return ret;
}
if (pkt && pkt->dts != AV_NOPTS_VALUE) {
ist->next_dts = ist->dts = av_rescale_q(pkt->dts, ist->st->time_base, AV_TIME_BASE_Q);
if (par->codec_type != AVMEDIA_TYPE_VIDEO)
ist->pts = ist->dts;
}
// while we have more to decode or while the decoder did output something on EOF
while (ist->decoding_needed) {
int64_t duration_dts = 0;
int64_t duration_pts = 0;
int got_output = 0;
int decode_failed = 0;
ist->pts = ist->next_pts;
ist->dts = ist->next_dts;
switch (par->codec_type) {
case AVMEDIA_TYPE_AUDIO:
ret = decode_audio (ist, repeating ? NULL : avpkt, &got_output,
&decode_failed);
av_packet_unref(avpkt);
break;
case AVMEDIA_TYPE_VIDEO:
ret = decode_video (ist, repeating ? NULL : avpkt, &got_output, &duration_pts, !pkt,
&decode_failed);
if (!repeating || !pkt || got_output) {
if (pkt && pkt->duration) {
duration_dts = av_rescale_q(pkt->duration, ist->st->time_base, AV_TIME_BASE_Q);
} else if(ist->dec_ctx->framerate.num != 0 && ist->dec_ctx->framerate.den != 0) {
int ticks = ist->last_pkt_repeat_pict >= 0 ?
ist->last_pkt_repeat_pict + 1 :
ist->dec_ctx->ticks_per_frame;
duration_dts = ((int64_t)AV_TIME_BASE *
ist->dec_ctx->framerate.den * ticks) /
ist->dec_ctx->framerate.num / ist->dec_ctx->ticks_per_frame;
}
if(ist->dts != AV_NOPTS_VALUE && duration_dts) {
ist->next_dts += duration_dts;
}else
ist->next_dts = AV_NOPTS_VALUE;
}
if (got_output) {
if (duration_pts > 0) {
ist->next_pts += av_rescale_q(duration_pts, ist->st->time_base, AV_TIME_BASE_Q);
} else {
ist->next_pts += duration_dts;
}
}
av_packet_unref(avpkt);
break;
case AVMEDIA_TYPE_SUBTITLE:
if (repeating)
break;
ret = transcode_subtitles(ist, avpkt, &got_output, &decode_failed);
if (!pkt && ret >= 0)
ret = AVERROR_EOF;
av_packet_unref(avpkt);
break;
2011-11-21 16:06:25 +03:00
default:
return -1;
}
if (ret == AVERROR_EOF) {
eof_reached = 1;
break;
}
if (ret < 0) {
if (decode_failed) {
av_log(NULL, AV_LOG_ERROR, "Error while decoding stream #%d:%d: %s\n",
ist->file_index, ist->st->index, av_err2str(ret));
} else {
av_log(NULL, AV_LOG_FATAL, "Error while processing the decoded "
"data for stream #%d:%d\n", ist->file_index, ist->st->index);
}
if (!decode_failed || exit_on_error)
exit_program(1);
break;
}
if (got_output)
ist->got_output = 1;
if (!got_output)
break;
// During draining, we might get multiple output frames in this loop.
// ffmpeg.c does not drain the filter chain on configuration changes,
// which means if we send multiple frames at once to the filters, and
// one of those frames changes configuration, the buffered frames will
// be lost. This can upset certain FATE tests.
// Decode only 1 frame per call on EOF to appease these FATE tests.
// The ideal solution would be to rewrite decoding to use the new
// decoding API in a better way.
if (!pkt)
break;
repeating = 1;
}
/* after flushing, send an EOF on all the filter inputs attached to the stream */
/* except when looping we need to flush but not to send an EOF */
if (!pkt && ist->decoding_needed && eof_reached && !no_eof) {
int ret = send_filter_eof(ist);
if (ret < 0) {
av_log(NULL, AV_LOG_FATAL, "Error marking filters as finished\n");
exit_program(1);
}
}
/* handle stream copy */
if (!ist->decoding_needed && pkt) {
ist->dts = ist->next_dts;
switch (par->codec_type) {
case AVMEDIA_TYPE_AUDIO:
av_assert1(pkt->duration >= 0);
if (par->sample_rate) {
ist->next_dts += ((int64_t)AV_TIME_BASE * par->frame_size) /
par->sample_rate;
} else {
ist->next_dts += av_rescale_q(pkt->duration, ist->st->time_base, AV_TIME_BASE_Q);
}
break;
case AVMEDIA_TYPE_VIDEO:
if (ist->framerate.num) {
// TODO: Remove work-around for c99-to-c89 issue 7
AVRational time_base_q = AV_TIME_BASE_Q;
int64_t next_dts = av_rescale_q(ist->next_dts, time_base_q, av_inv_q(ist->framerate));
ist->next_dts = av_rescale_q(next_dts + 1, av_inv_q(ist->framerate), time_base_q);
} else if (pkt->duration) {
ist->next_dts += av_rescale_q(pkt->duration, ist->st->time_base, AV_TIME_BASE_Q);
} else if(ist->dec_ctx->framerate.num != 0) {
int ticks = ist->last_pkt_repeat_pict >= 0 ?
ist->last_pkt_repeat_pict + 1 :
ist->dec_ctx->ticks_per_frame;
ist->next_dts += ((int64_t)AV_TIME_BASE *
ist->dec_ctx->framerate.den * ticks) /
ist->dec_ctx->framerate.num / ist->dec_ctx->ticks_per_frame;
}
break;
}
ist->pts = ist->dts;
} else if (!ist->decoding_needed)
eof_reached = 1;
duration_exceeded = 0;
if (f->recording_time != INT64_MAX) {
int64_t start_time = 0;
if (copy_ts) {
start_time += f->start_time != AV_NOPTS_VALUE ? f->start_time : 0;
start_time += start_at_zero ? 0 : f->start_time_effective;
}
if (ist->dts >= f->recording_time + start_time)
duration_exceeded = 1;
}
for (int oidx = 0; oidx < ist->nb_outputs; oidx++) {
OutputStream *ost = ist->outputs[oidx];
if (ost->enc_ctx || (!pkt && no_eof))
continue;
if (duration_exceeded) {
close_output_stream(ost);
continue;
}
do_streamcopy(ist, ost, pkt, ist->dts);
}
return !eof_reached;
}
static enum AVPixelFormat get_format(AVCodecContext *s, const enum AVPixelFormat *pix_fmts)
{
InputStream *ist = s->opaque;
const enum AVPixelFormat *p;
int ret;
for (p = pix_fmts; *p != AV_PIX_FMT_NONE; p++) {
const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(*p);
const AVCodecHWConfig *config = NULL;
int i;
if (!(desc->flags & AV_PIX_FMT_FLAG_HWACCEL))
break;
if (ist->hwaccel_id == HWACCEL_GENERIC ||
ist->hwaccel_id == HWACCEL_AUTO) {
for (i = 0;; i++) {
config = avcodec_get_hw_config(s->codec, i);
if (!config)
break;
if (!(config->methods &
AV_CODEC_HW_CONFIG_METHOD_HW_DEVICE_CTX))
continue;
if (config->pix_fmt == *p)
break;
}
}
if (config && config->device_type == ist->hwaccel_device_type) {
ret = hwaccel_decode_init(s);
if (ret < 0) {
if (ist->hwaccel_id == HWACCEL_GENERIC) {
av_log(NULL, AV_LOG_FATAL,
"%s hwaccel requested for input stream #%d:%d, "
"but cannot be initialized.\n",
av_hwdevice_get_type_name(config->device_type),
ist->file_index, ist->st->index);
return AV_PIX_FMT_NONE;
}
continue;
}
ist->hwaccel_pix_fmt = *p;
break;
}
}
return *p;
}
static int init_input_stream(InputStream *ist, char *error, int error_len)
{
int ret;
if (ist->decoding_needed) {
const AVCodec *codec = ist->dec;
if (!codec) {
snprintf(error, error_len, "Decoder (codec %s) not found for input stream #%d:%d",
avcodec_get_name(ist->dec_ctx->codec_id), ist->file_index, ist->st->index);
return AVERROR(EINVAL);
}
ist->dec_ctx->opaque = ist;
ist->dec_ctx->get_format = get_format;
if (ist->dec_ctx->codec_id == AV_CODEC_ID_DVB_SUBTITLE &&
(ist->decoding_needed & DECODING_FOR_OST)) {
av_dict_set(&ist->decoder_opts, "compute_edt", "1", AV_DICT_DONT_OVERWRITE);
if (ist->decoding_needed & DECODING_FOR_FILTER)
av_log(NULL, AV_LOG_WARNING, "Warning using DVB subtitles for filtering and output at the same time is not fully supported, also see -compute_edt [0|1]\n");
}
/* Useful for subtitles retiming by lavf (FIXME), skipping samples in
* audio, and video decoders such as cuvid or mediacodec */
ist->dec_ctx->pkt_timebase = ist->st->time_base;
if (!av_dict_get(ist->decoder_opts, "threads", NULL, 0))
av_dict_set(&ist->decoder_opts, "threads", "auto", 0);
/* Attached pics are sparse, therefore we would not want to delay their decoding till EOF. */
if (ist->st->disposition & AV_DISPOSITION_ATTACHED_PIC)
av_dict_set(&ist->decoder_opts, "threads", "1", 0);
ret = hw_device_setup_for_decode(ist);
if (ret < 0) {
snprintf(error, error_len, "Device setup failed for "
"decoder on input stream #%d:%d : %s",
ist->file_index, ist->st->index, av_err2str(ret));
return ret;
}
if ((ret = avcodec_open2(ist->dec_ctx, codec, &ist->decoder_opts)) < 0) {
if (ret == AVERROR_EXPERIMENTAL)
abort_codec_experimental(codec, 0);
snprintf(error, error_len,
"Error while opening decoder for input stream "
"#%d:%d : %s",
ist->file_index, ist->st->index, av_err2str(ret));
return ret;
}
assert_avoptions(ist->decoder_opts);
}
return 0;
}
static int init_output_stream_streamcopy(OutputStream *ost)
{
OutputFile *of = output_files[ost->file_index];
InputStream *ist = ost->ist;
InputFile *ifile = input_files[ist->file_index];
AVCodecParameters *par = ost->st->codecpar;
AVCodecContext *codec_ctx;
AVRational sar;
int i, ret;
uint32_t codec_tag = par->codec_tag;
av_assert0(ist && !ost->filter);
codec_ctx = avcodec_alloc_context3(NULL);
if (!codec_ctx)
return AVERROR(ENOMEM);
ret = avcodec_parameters_to_context(codec_ctx, ist->par);
if (ret >= 0)
ret = av_opt_set_dict(codec_ctx, &ost->encoder_opts);
if (ret < 0) {
av_log(ost, AV_LOG_FATAL,
"Error setting up codec context options.\n");
avcodec_free_context(&codec_ctx);
return ret;
}
ret = avcodec_parameters_from_context(par, codec_ctx);
avcodec_free_context(&codec_ctx);
if (ret < 0) {
av_log(ost, AV_LOG_FATAL,
"Error getting reference codec parameters.\n");
return ret;
}
if (!codec_tag) {
unsigned int codec_tag_tmp;
if (!of->format->codec_tag ||
av_codec_get_id (of->format->codec_tag, par->codec_tag) == par->codec_id ||
!av_codec_get_tag2(of->format->codec_tag, par->codec_id, &codec_tag_tmp))
codec_tag = par->codec_tag;
}
par->codec_tag = codec_tag;
if (!ost->frame_rate.num)
ost->frame_rate = ist->framerate;
if (ost->frame_rate.num)
ost->st->avg_frame_rate = ost->frame_rate;
else
ost->st->avg_frame_rate = ist->st->avg_frame_rate;
ret = avformat_transfer_internal_stream_timing_info(of->format, ost->st, ist->st, copy_tb);
if (ret < 0)
return ret;
// copy timebase while removing common factors
if (ost->st->time_base.num <= 0 || ost->st->time_base.den <= 0) {
if (ost->frame_rate.num)
ost->st->time_base = av_inv_q(ost->frame_rate);
else
ost->st->time_base = av_add_q(av_stream_get_codec_timebase(ost->st), (AVRational){0, 1});
}
// copy estimated duration as a hint to the muxer
if (ost->st->duration <= 0 && ist->st->duration > 0)
ost->st->duration = av_rescale_q(ist->st->duration, ist->st->time_base, ost->st->time_base);
if (!ost->copy_prior_start) {
ost->ts_copy_start = (of->start_time == AV_NOPTS_VALUE) ?
0 : of->start_time;
if (copy_ts && ifile->start_time != AV_NOPTS_VALUE) {
ost->ts_copy_start = FFMAX(ost->ts_copy_start,
ifile->start_time + ifile->ts_offset);
}
}
if (ist->st->nb_side_data) {
for (i = 0; i < ist->st->nb_side_data; i++) {
const AVPacketSideData *sd_src = &ist->st->side_data[i];
uint8_t *dst_data;
dst_data = av_stream_new_side_data(ost->st, sd_src->type, sd_src->size);
if (!dst_data)
return AVERROR(ENOMEM);
memcpy(dst_data, sd_src->data, sd_src->size);
}
}
#if FFMPEG_ROTATION_METADATA
avformat, ffmpeg: deprecate old rotation API The old "API" that signaled rotation as a metadata value has been replaced by DISPLAYMATRIX side data quite a while ago. There is no reason to make muxers/demuxers/API users support both. In addition, the metadata API is dangerous, as user tags could "leak" into it, creating unintended features or bugs. ffmpeg CLI has to be updated to use the new API. In particular, we must not allow to leak the "rotate" tag into the muxer. Some muxers will catch this properly (like mov), but others (like mkv) can add it as generic tag. Note applications, which use libavformat and assume the old rotate API, will interpret such "rotate" user tags as rotate metadata (which it is not), and incorrectly rotate the video. The ffmpeg/ffplay tools drop the use of the old API for muxing and demuxing, as all muxers/demuxers support the new API. This will mean that the tools will not mistakenly interpret per-track "rotate" user tags as rotate metadata. It will _not_ be treated as regression. Unfortunately, hacks have been added, that allow the user to override rotation by setting metadata explicitly, e.g. via -metadata:s:v:0 rotate=0 See references to trac #4560. fate-filter-meta-4560-rotate0 tests this. It's easier to adjust the hack for supporting it than arguing for its removal, so ffmpeg CLI now explicitly catches this case, and essentially replaces the "rotate" value with a display matrix side data. (It would be easier for both user and implementation to create an explicit option for rotation.) When the code under FF_API_OLD_ROTATE_API is disabled, one FATE reference file has to be updated (because "rotate" is not exported anymore). Tested-by: Michael Niedermayer <michael@niedermayer.cc> Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
2017-03-21 09:02:58 +02:00
if (ost->rotate_overridden) {
uint8_t *sd = av_stream_new_side_data(ost->st, AV_PKT_DATA_DISPLAYMATRIX,
sizeof(int32_t) * 9);
if (sd)
av_display_rotation_set((int32_t *)sd, -ost->rotate_override_value);
}
#endif
avformat, ffmpeg: deprecate old rotation API The old "API" that signaled rotation as a metadata value has been replaced by DISPLAYMATRIX side data quite a while ago. There is no reason to make muxers/demuxers/API users support both. In addition, the metadata API is dangerous, as user tags could "leak" into it, creating unintended features or bugs. ffmpeg CLI has to be updated to use the new API. In particular, we must not allow to leak the "rotate" tag into the muxer. Some muxers will catch this properly (like mov), but others (like mkv) can add it as generic tag. Note applications, which use libavformat and assume the old rotate API, will interpret such "rotate" user tags as rotate metadata (which it is not), and incorrectly rotate the video. The ffmpeg/ffplay tools drop the use of the old API for muxing and demuxing, as all muxers/demuxers support the new API. This will mean that the tools will not mistakenly interpret per-track "rotate" user tags as rotate metadata. It will _not_ be treated as regression. Unfortunately, hacks have been added, that allow the user to override rotation by setting metadata explicitly, e.g. via -metadata:s:v:0 rotate=0 See references to trac #4560. fate-filter-meta-4560-rotate0 tests this. It's easier to adjust the hack for supporting it than arguing for its removal, so ffmpeg CLI now explicitly catches this case, and essentially replaces the "rotate" value with a display matrix side data. (It would be easier for both user and implementation to create an explicit option for rotation.) When the code under FF_API_OLD_ROTATE_API is disabled, one FATE reference file has to be updated (because "rotate" is not exported anymore). Tested-by: Michael Niedermayer <michael@niedermayer.cc> Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
2017-03-21 09:02:58 +02:00
switch (par->codec_type) {
case AVMEDIA_TYPE_AUDIO:
if ((par->block_align == 1 || par->block_align == 1152 || par->block_align == 576) &&
par->codec_id == AV_CODEC_ID_MP3)
par->block_align = 0;
if (par->codec_id == AV_CODEC_ID_AC3)
par->block_align = 0;
break;
case AVMEDIA_TYPE_VIDEO:
if (ost->frame_aspect_ratio.num) { // overridden by the -aspect cli option
sar =
av_mul_q(ost->frame_aspect_ratio,
(AVRational){ par->height, par->width });
av_log(ost, AV_LOG_WARNING, "Overriding aspect ratio "
"with stream copy may produce invalid files\n");
}
else if (ist->st->sample_aspect_ratio.num)
sar = ist->st->sample_aspect_ratio;
else
sar = par->sample_aspect_ratio;
ost->st->sample_aspect_ratio = par->sample_aspect_ratio = sar;
ost->st->avg_frame_rate = ist->st->avg_frame_rate;
ost->st->r_frame_rate = ist->st->r_frame_rate;
break;
}
ost->mux_timebase = ist->st->time_base;
return 0;
}
static int init_output_stream_nofilter(OutputStream *ost)
{
int ret = 0;
if (ost->enc_ctx) {
ret = enc_open(ost, NULL);
if (ret < 0)
return ret;
} else {
if (ost->ist) {
ret = init_output_stream_streamcopy(ost);
if (ret < 0)
return ret;
}
ret = of_stream_init(output_files[ost->file_index], ost);
if (ret < 0)
return ret;
}
return ret;
}
static int transcode_init(void)
{
int ret = 0;
char error[1024] = {0};
/* init framerate emulation */
for (int i = 0; i < nb_input_files; i++) {
InputFile *ifile = input_files[i];
if (ifile->readrate || ifile->rate_emu)
for (int j = 0; j < ifile->nb_streams; j++)
ifile->streams[j]->start = av_gettime_relative();
}
/* init input streams */
for (InputStream *ist = ist_iter(NULL); ist; ist = ist_iter(ist))
if ((ret = init_input_stream(ist, error, sizeof(error))) < 0)
goto dump_format;
/*
* initialize stream copy and subtitle/data streams.
* Encoded AVFrame based streams will get initialized when the first AVFrame
* is received in do_video_out
*/
for (OutputStream *ost = ost_iter(NULL); ost; ost = ost_iter(ost)) {
if (ost->enc_ctx &&
(ost->st->codecpar->codec_type == AVMEDIA_TYPE_VIDEO ||
ost->st->codecpar->codec_type == AVMEDIA_TYPE_AUDIO))
continue;
ret = init_output_stream_nofilter(ost);
if (ret < 0)
goto dump_format;
}
/* discard unused programs */
for (int i = 0; i < nb_input_files; i++) {
InputFile *ifile = input_files[i];
for (int j = 0; j < ifile->ctx->nb_programs; j++) {
AVProgram *p = ifile->ctx->programs[j];
int discard = AVDISCARD_ALL;
for (int k = 0; k < p->nb_stream_indexes; k++)
if (!ifile->streams[p->stream_index[k]]->discard) {
discard = AVDISCARD_DEFAULT;
break;
}
p->discard = discard;
}
}
dump_format:
/* dump the stream mapping */
av_log(NULL, AV_LOG_INFO, "Stream mapping:\n");
for (InputStream *ist = ist_iter(NULL); ist; ist = ist_iter(ist)) {
for (int j = 0; j < ist->nb_filters; j++) {
if (!filtergraph_is_simple(ist->filters[j]->graph)) {
av_log(NULL, AV_LOG_INFO, " Stream #%d:%d (%s) -> %s",
ist->file_index, ist->st->index, ist->dec ? ist->dec->name : "?",
ist->filters[j]->name);
if (nb_filtergraphs > 1)
av_log(NULL, AV_LOG_INFO, " (graph %d)", ist->filters[j]->graph->index);
av_log(NULL, AV_LOG_INFO, "\n");
}
}
}
for (OutputStream *ost = ost_iter(NULL); ost; ost = ost_iter(ost)) {
if (ost->attachment_filename) {
/* an attached file */
av_log(NULL, AV_LOG_INFO, " File %s -> Stream #%d:%d\n",
ost->attachment_filename, ost->file_index, ost->index);
continue;
}
if (ost->filter && !filtergraph_is_simple(ost->filter->graph)) {
/* output from a complex graph */
av_log(NULL, AV_LOG_INFO, " %s", ost->filter->name);
if (nb_filtergraphs > 1)
av_log(NULL, AV_LOG_INFO, " (graph %d)", ost->filter->graph->index);
av_log(NULL, AV_LOG_INFO, " -> Stream #%d:%d (%s)\n", ost->file_index,
ost->index, ost->enc_ctx->codec->name);
continue;
}
av_log(NULL, AV_LOG_INFO, " Stream #%d:%d -> #%d:%d",
ost->ist->file_index,
ost->ist->st->index,
ost->file_index,
ost->index);
if (ost->enc_ctx) {
const AVCodec *in_codec = ost->ist->dec;
const AVCodec *out_codec = ost->enc_ctx->codec;
const char *decoder_name = "?";
const char *in_codec_name = "?";
const char *encoder_name = "?";
const char *out_codec_name = "?";
const AVCodecDescriptor *desc;
if (in_codec) {
decoder_name = in_codec->name;
desc = avcodec_descriptor_get(in_codec->id);
if (desc)
in_codec_name = desc->name;
if (!strcmp(decoder_name, in_codec_name))
decoder_name = "native";
}
if (out_codec) {
encoder_name = out_codec->name;
desc = avcodec_descriptor_get(out_codec->id);
if (desc)
out_codec_name = desc->name;
if (!strcmp(encoder_name, out_codec_name))
encoder_name = "native";
}
av_log(NULL, AV_LOG_INFO, " (%s (%s) -> %s (%s))",
in_codec_name, decoder_name,
out_codec_name, encoder_name);
} else
av_log(NULL, AV_LOG_INFO, " (copy)");
av_log(NULL, AV_LOG_INFO, "\n");
}
if (ret) {
av_log(NULL, AV_LOG_ERROR, "%s\n", error);
return ret;
}
atomic_store(&transcode_init_done, 1);
return 0;
}
2012-08-24 14:31:50 +03:00
/* Return 1 if there remain streams where more output is wanted, 0 otherwise. */
static int need_output(void)
{
for (OutputStream *ost = ost_iter(NULL); ost; ost = ost_iter(ost)) {
if (ost->finished)
continue;
return 1;
}
return 0;
}
/**
* Select the output stream to process.
*
* @return selected output stream, or NULL if none available
*/
static OutputStream *choose_output(void)
{
int64_t opts_min = INT64_MAX;
OutputStream *ost_min = NULL;
for (OutputStream *ost = ost_iter(NULL); ost; ost = ost_iter(ost)) {
int64_t opts;
if (ost->filter && ost->filter->last_pts != AV_NOPTS_VALUE) {
opts = ost->filter->last_pts;
} else {
opts = ost->last_mux_dts == AV_NOPTS_VALUE ?
INT64_MIN : ost->last_mux_dts;
if (ost->last_mux_dts == AV_NOPTS_VALUE)
av_log(ost, AV_LOG_DEBUG,
"cur_dts is invalid [init:%d i_done:%d finish:%d] (this is harmless if it occurs once at the start per stream)\n",
ost->initialized, ost->inputs_done, ost->finished);
}
if (!ost->initialized && !ost->inputs_done && !ost->finished)
return ost->unavailable ? NULL : ost;
if (!ost->finished && opts < opts_min) {
opts_min = opts;
ost_min = ost->unavailable ? NULL : ost;
}
}
return ost_min;
}
static void set_tty_echo(int on)
{
#if HAVE_TERMIOS_H
struct termios tty;
if (tcgetattr(0, &tty) == 0) {
if (on) tty.c_lflag |= ECHO;
else tty.c_lflag &= ~ECHO;
tcsetattr(0, TCSANOW, &tty);
}
#endif
}
static int check_keyboard_interaction(int64_t cur_time)
{
int i, ret, key;
static int64_t last_time;
if (received_nb_signals)
return AVERROR_EXIT;
/* read_key() returns 0 on EOF */
if (cur_time - last_time >= 100000) {
key = read_key();
last_time = cur_time;
}else
key = -1;
if (key == 'q') {
av_log(NULL, AV_LOG_INFO, "\n\n[q] command received. Exiting.\n\n");
return AVERROR_EXIT;
}
if (key == '+') av_log_set_level(av_log_get_level()+10);
if (key == '-') av_log_set_level(av_log_get_level()-10);
if (key == 'c' || key == 'C'){
char buf[4096], target[64], command[256], arg[256] = {0};
double time;
int k, n = 0;
fprintf(stderr, "\nEnter command: <target>|all <time>|-1 <command>[ <argument>]\n");
i = 0;
set_tty_echo(1);
while ((k = read_key()) != '\n' && k != '\r' && i < sizeof(buf)-1)
if (k > 0)
buf[i++] = k;
buf[i] = 0;
set_tty_echo(0);
fprintf(stderr, "\n");
if (k > 0 &&
(n = sscanf(buf, "%63[^ ] %lf %255[^ ] %255[^\n]", target, &time, command, arg)) >= 3) {
av_log(NULL, AV_LOG_DEBUG, "Processing command target:%s time:%f command:%s arg:%s",
target, time, command, arg);
for (i = 0; i < nb_filtergraphs; i++) {
FilterGraph *fg = filtergraphs[i];
if (fg->graph) {
if (time < 0) {
ret = avfilter_graph_send_command(fg->graph, target, command, arg, buf, sizeof(buf),
key == 'c' ? AVFILTER_CMD_FLAG_ONE : 0);
fprintf(stderr, "Command reply for stream %d: ret:%d res:\n%s", i, ret, buf);
} else if (key == 'c') {
fprintf(stderr, "Queuing commands only on filters supporting the specific command is unsupported\n");
ret = AVERROR_PATCHWELCOME;
} else {
ret = avfilter_graph_queue_command(fg->graph, target, command, arg, 0, time);
if (ret < 0)
fprintf(stderr, "Queuing command failed with error %s\n", av_err2str(ret));
}
}
}
} else {
av_log(NULL, AV_LOG_ERROR,
"Parse error, at least 3 arguments were expected, "
"only %d given in string '%s'\n", n, buf);
}
}
if (key == 'd' || key == 'D'){
int debug=0;
if(key == 'D') {
InputStream *ist = ist_iter(NULL);
if (ist)
debug = ist->dec_ctx->debug << 1;
if(!debug) debug = 1;
while (debug & FF_DEBUG_DCT_COEFF) //unsupported, would just crash
debug += debug;
}else{
char buf[32];
int k = 0;
i = 0;
set_tty_echo(1);
while ((k = read_key()) != '\n' && k != '\r' && i < sizeof(buf)-1)
if (k > 0)
buf[i++] = k;
buf[i] = 0;
set_tty_echo(0);
fprintf(stderr, "\n");
if (k <= 0 || sscanf(buf, "%d", &debug)!=1)
fprintf(stderr,"error parsing debug value\n");
}
for (InputStream *ist = ist_iter(NULL); ist; ist = ist_iter(ist))
ist->dec_ctx->debug = debug;
for (OutputStream *ost = ost_iter(NULL); ost; ost = ost_iter(ost)) {
if (ost->enc_ctx)
ost->enc_ctx->debug = debug;
}
if(debug) av_log_set_level(AV_LOG_DEBUG);
fprintf(stderr,"debug=%d\n", debug);
}
if (key == '?'){
fprintf(stderr, "key function\n"
"? show this help\n"
"+ increase verbosity\n"
"- decrease verbosity\n"
"c Send command to first matching filter supporting it\n"
"C Send/Queue command to all matching filters\n"
"D cycle through available debug modes\n"
"h dump packets/hex press to cycle through the 3 states\n"
"q quit\n"
"s Show QP histogram\n"
);
}
return 0;
}
static int got_eagain(void)
{
for (OutputStream *ost = ost_iter(NULL); ost; ost = ost_iter(ost))
if (ost->unavailable)
return 1;
return 0;
}
static void reset_eagain(void)
{
int i;
for (i = 0; i < nb_input_files; i++)
input_files[i]->eagain = 0;
for (OutputStream *ost = ost_iter(NULL); ost; ost = ost_iter(ost))
ost->unavailable = 0;
}
static void decode_flush(InputFile *ifile)
{
for (int i = 0; i < ifile->nb_streams; i++) {
InputStream *ist = ifile->streams[i];
int ret;
if (ist->discard)
continue;
do {
ret = process_input_packet(ist, NULL, 1);
} while (ret > 0);
if (ist->decoding_needed) {
/* report last frame duration to the demuxer thread */
if (ist->par->codec_type == AVMEDIA_TYPE_AUDIO) {
LastFrameDuration dur;
dur.stream_idx = i;
dur.duration = av_rescale_q(ist->nb_samples,
(AVRational){ 1, ist->dec_ctx->sample_rate},
ist->st->time_base);
av_thread_message_queue_send(ifile->audio_duration_queue, &dur, 0);
}
avcodec_flush_buffers(ist->dec_ctx);
}
}
}
static void ts_discontinuity_detect(InputFile *ifile, InputStream *ist,
AVPacket *pkt)
{
const int fmt_is_discont = ifile->ctx->iformat->flags & AVFMT_TS_DISCONT;
int disable_discontinuity_correction = copy_ts;
int64_t pkt_dts = av_rescale_q_rnd(pkt->dts, ist->st->time_base, AV_TIME_BASE_Q,
AV_ROUND_NEAR_INF | AV_ROUND_PASS_MINMAX);
if (copy_ts && ist->next_dts != AV_NOPTS_VALUE &&
fmt_is_discont && ist->st->pts_wrap_bits < 60) {
int64_t wrap_dts = av_rescale_q_rnd(pkt->dts + (1LL<<ist->st->pts_wrap_bits),
ist->st->time_base, AV_TIME_BASE_Q,
AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
if (FFABS(wrap_dts - ist->next_dts) < FFABS(pkt_dts - ist->next_dts)/10)
disable_discontinuity_correction = 0;
}
if (ist->next_dts != AV_NOPTS_VALUE && !disable_discontinuity_correction) {
int64_t delta = pkt_dts - ist->next_dts;
if (fmt_is_discont) {
if (FFABS(delta) > 1LL * dts_delta_threshold * AV_TIME_BASE ||
pkt_dts + AV_TIME_BASE/10 < FFMAX(ist->pts, ist->dts)) {
ifile->ts_offset_discont -= delta;
av_log(NULL, AV_LOG_DEBUG,
"timestamp discontinuity for stream #%d:%d "
"(id=%d, type=%s): %"PRId64", new offset= %"PRId64"\n",
ist->file_index, ist->st->index, ist->st->id,
av_get_media_type_string(ist->par->codec_type),
delta, ifile->ts_offset_discont);
pkt->dts -= av_rescale_q(delta, AV_TIME_BASE_Q, ist->st->time_base);
if (pkt->pts != AV_NOPTS_VALUE)
pkt->pts -= av_rescale_q(delta, AV_TIME_BASE_Q, ist->st->time_base);
}
} else {
if (FFABS(delta) > 1LL * dts_error_threshold * AV_TIME_BASE) {
av_log(NULL, AV_LOG_WARNING, "DTS %"PRId64", next:%"PRId64" st:%d invalid dropping\n", pkt->dts, ist->next_dts, pkt->stream_index);
pkt->dts = AV_NOPTS_VALUE;
}
if (pkt->pts != AV_NOPTS_VALUE){
int64_t pkt_pts = av_rescale_q(pkt->pts, ist->st->time_base, AV_TIME_BASE_Q);
delta = pkt_pts - ist->next_dts;
if (FFABS(delta) > 1LL * dts_error_threshold * AV_TIME_BASE) {
av_log(NULL, AV_LOG_WARNING, "PTS %"PRId64", next:%"PRId64" invalid dropping st:%d\n", pkt->pts, ist->next_dts, pkt->stream_index);
pkt->pts = AV_NOPTS_VALUE;
}
}
}
} else if (ist->next_dts == AV_NOPTS_VALUE && !copy_ts &&
fmt_is_discont && ifile->last_ts != AV_NOPTS_VALUE) {
int64_t delta = pkt_dts - ifile->last_ts;
if (FFABS(delta) > 1LL * dts_delta_threshold * AV_TIME_BASE) {
ifile->ts_offset_discont -= delta;
av_log(NULL, AV_LOG_DEBUG,
"Inter stream timestamp discontinuity %"PRId64", new offset= %"PRId64"\n",
delta, ifile->ts_offset_discont);
pkt->dts -= av_rescale_q(delta, AV_TIME_BASE_Q, ist->st->time_base);
if (pkt->pts != AV_NOPTS_VALUE)
pkt->pts -= av_rescale_q(delta, AV_TIME_BASE_Q, ist->st->time_base);
}
}
ifile->last_ts = av_rescale_q(pkt->dts, ist->st->time_base, AV_TIME_BASE_Q);
}
static void ts_discontinuity_process(InputFile *ifile, InputStream *ist,
AVPacket *pkt)
{
int64_t offset = av_rescale_q(ifile->ts_offset_discont, AV_TIME_BASE_Q,
ist->st->time_base);
// apply previously-detected timestamp-discontinuity offset
// (to all streams, not just audio/video)
if (pkt->dts != AV_NOPTS_VALUE)
pkt->dts += offset;
if (pkt->pts != AV_NOPTS_VALUE)
pkt->pts += offset;
// detect timestamp discontinuities for audio/video
if ((ist->par->codec_type == AVMEDIA_TYPE_VIDEO ||
ist->par->codec_type == AVMEDIA_TYPE_AUDIO) &&
pkt->dts != AV_NOPTS_VALUE)
ts_discontinuity_detect(ifile, ist, pkt);
}
2012-08-24 14:31:50 +03:00
/*
* Return
* - 0 -- one packet was read and processed
* - AVERROR(EAGAIN) -- no packets were available for selected file,
* this function should be called again
* - AVERROR_EOF -- this function should not be called again
*/
static int process_input(int file_index)
{
InputFile *ifile = input_files[file_index];
AVFormatContext *is;
InputStream *ist;
AVPacket *pkt;
int ret, i;
is = ifile->ctx;
ret = ifile_get_packet(ifile, &pkt);
if (ret == AVERROR(EAGAIN)) {
ifile->eagain = 1;
return ret;
}
if (ret == 1) {
/* the input file is looped: flush the decoders */
decode_flush(ifile);
return AVERROR(EAGAIN);
}
if (ret < 0) {
if (ret != AVERROR_EOF) {
print_error(is->url, ret);
if (exit_on_error)
exit_program(1);
}
for (i = 0; i < ifile->nb_streams; i++) {
ist = ifile->streams[i];
if (!ist->discard) {
ret = process_input_packet(ist, NULL, 0);
if (ret>0)
return 0;
}
/* mark all outputs that don't go through lavfi as finished */
for (int oidx = 0; oidx < ist->nb_outputs; oidx++) {
OutputStream *ost = ist->outputs[oidx];
OutputFile *of = output_files[ost->file_index];
close_output_stream(ost);
of_output_packet(of, ost->pkt, ost, 1);
}
}
ifile->eof_reached = 1;
return AVERROR(EAGAIN);
}
reset_eagain();
ist = ifile->streams[pkt->stream_index];
ist->data_size += pkt->size;
ist->nb_packets++;
if (ist->discard)
goto discard_packet;
/* add the stream-global side data to the first packet */
if (ist->nb_packets == 1) {
for (i = 0; i < ist->st->nb_side_data; i++) {
AVPacketSideData *src_sd = &ist->st->side_data[i];
uint8_t *dst_data;
avformat, ffmpeg: deprecate old rotation API The old "API" that signaled rotation as a metadata value has been replaced by DISPLAYMATRIX side data quite a while ago. There is no reason to make muxers/demuxers/API users support both. In addition, the metadata API is dangerous, as user tags could "leak" into it, creating unintended features or bugs. ffmpeg CLI has to be updated to use the new API. In particular, we must not allow to leak the "rotate" tag into the muxer. Some muxers will catch this properly (like mov), but others (like mkv) can add it as generic tag. Note applications, which use libavformat and assume the old rotate API, will interpret such "rotate" user tags as rotate metadata (which it is not), and incorrectly rotate the video. The ffmpeg/ffplay tools drop the use of the old API for muxing and demuxing, as all muxers/demuxers support the new API. This will mean that the tools will not mistakenly interpret per-track "rotate" user tags as rotate metadata. It will _not_ be treated as regression. Unfortunately, hacks have been added, that allow the user to override rotation by setting metadata explicitly, e.g. via -metadata:s:v:0 rotate=0 See references to trac #4560. fate-filter-meta-4560-rotate0 tests this. It's easier to adjust the hack for supporting it than arguing for its removal, so ffmpeg CLI now explicitly catches this case, and essentially replaces the "rotate" value with a display matrix side data. (It would be easier for both user and implementation to create an explicit option for rotation.) When the code under FF_API_OLD_ROTATE_API is disabled, one FATE reference file has to be updated (because "rotate" is not exported anymore). Tested-by: Michael Niedermayer <michael@niedermayer.cc> Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
2017-03-21 09:02:58 +02:00
if (src_sd->type == AV_PKT_DATA_DISPLAYMATRIX)
continue;
avformat, ffmpeg: deprecate old rotation API The old "API" that signaled rotation as a metadata value has been replaced by DISPLAYMATRIX side data quite a while ago. There is no reason to make muxers/demuxers/API users support both. In addition, the metadata API is dangerous, as user tags could "leak" into it, creating unintended features or bugs. ffmpeg CLI has to be updated to use the new API. In particular, we must not allow to leak the "rotate" tag into the muxer. Some muxers will catch this properly (like mov), but others (like mkv) can add it as generic tag. Note applications, which use libavformat and assume the old rotate API, will interpret such "rotate" user tags as rotate metadata (which it is not), and incorrectly rotate the video. The ffmpeg/ffplay tools drop the use of the old API for muxing and demuxing, as all muxers/demuxers support the new API. This will mean that the tools will not mistakenly interpret per-track "rotate" user tags as rotate metadata. It will _not_ be treated as regression. Unfortunately, hacks have been added, that allow the user to override rotation by setting metadata explicitly, e.g. via -metadata:s:v:0 rotate=0 See references to trac #4560. fate-filter-meta-4560-rotate0 tests this. It's easier to adjust the hack for supporting it than arguing for its removal, so ffmpeg CLI now explicitly catches this case, and essentially replaces the "rotate" value with a display matrix side data. (It would be easier for both user and implementation to create an explicit option for rotation.) When the code under FF_API_OLD_ROTATE_API is disabled, one FATE reference file has to be updated (because "rotate" is not exported anymore). Tested-by: Michael Niedermayer <michael@niedermayer.cc> Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
2017-03-21 09:02:58 +02:00
if (av_packet_get_side_data(pkt, src_sd->type, NULL))
continue;
dst_data = av_packet_new_side_data(pkt, src_sd->type, src_sd->size);
if (!dst_data)
report_and_exit(AVERROR(ENOMEM));
memcpy(dst_data, src_sd->data, src_sd->size);
}
}
// detect and try to correct for timestamp discontinuities
ts_discontinuity_process(ifile, ist, pkt);
if (debug_ts) {
av_log(NULL, AV_LOG_INFO, "demuxer+ffmpeg -> ist_index:%d:%d type:%s pkt_pts:%s pkt_pts_time:%s pkt_dts:%s pkt_dts_time:%s duration:%s duration_time:%s off:%s off_time:%s\n",
ifile->index, pkt->stream_index,
av_get_media_type_string(ist->par->codec_type),
av_ts2str(pkt->pts), av_ts2timestr(pkt->pts, &ist->st->time_base),
av_ts2str(pkt->dts), av_ts2timestr(pkt->dts, &ist->st->time_base),
av_ts2str(pkt->duration), av_ts2timestr(pkt->duration, &ist->st->time_base),
2012-11-22 02:12:18 +03:00
av_ts2str(input_files[ist->file_index]->ts_offset),
av_ts2timestr(input_files[ist->file_index]->ts_offset, &AV_TIME_BASE_Q));
}
sub2video_heartbeat(ist, pkt->pts);
process_input_packet(ist, pkt, 0);
discard_packet:
av_packet_free(&pkt);
return 0;
}
/**
* Perform a step of transcoding for the specified filter graph.
*
* @param[in] graph filter graph to consider
* @param[out] best_ist input stream where a frame would allow to continue
* @return 0 for success, <0 for error
*/
static int transcode_from_filter(FilterGraph *graph, InputStream **best_ist)
{
int i, ret;
int nb_requests, nb_requests_max = 0;
InputFilter *ifilter;
InputStream *ist;
*best_ist = NULL;
ret = avfilter_graph_request_oldest(graph->graph);
if (ret >= 0)
return reap_filters(0);
if (ret == AVERROR_EOF) {
ret = reap_filters(1);
for (i = 0; i < graph->nb_outputs; i++)
close_output_stream(graph->outputs[i]->ost);
return ret;
}
if (ret != AVERROR(EAGAIN))
return ret;
for (i = 0; i < graph->nb_inputs; i++) {
ifilter = graph->inputs[i];
ist = ifilter->ist;
if (input_files[ist->file_index]->eagain ||
input_files[ist->file_index]->eof_reached)
continue;
nb_requests = av_buffersrc_get_nb_failed_requests(ifilter->filter);
if (nb_requests > nb_requests_max) {
nb_requests_max = nb_requests;
*best_ist = ist;
}
}
if (!*best_ist)
for (i = 0; i < graph->nb_outputs; i++)
graph->outputs[i]->ost->unavailable = 1;
return 0;
}
/**
* Run a single step of transcoding.
*
* @return 0 for success, <0 for error
*/
static int transcode_step(void)
{
OutputStream *ost;
InputStream *ist = NULL;
int ret;
ost = choose_output();
if (!ost) {
if (got_eagain()) {
reset_eagain();
av_usleep(10000);
return 0;
}
av_log(NULL, AV_LOG_VERBOSE, "No more inputs to read from, finishing.\n");
return AVERROR_EOF;
}
if (ost->filter && !ost->filter->graph->graph) {
if (ifilter_has_all_input_formats(ost->filter->graph)) {
ret = configure_filtergraph(ost->filter->graph);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Error reinitializing filters!\n");
return ret;
}
}
}
if (ost->filter && ost->filter->graph->graph) {
if ((ret = transcode_from_filter(ost->filter->graph, &ist)) < 0)
return ret;
if (!ist)
return 0;
} else if (ost->filter) {
int i;
for (i = 0; i < ost->filter->graph->nb_inputs; i++) {
InputFilter *ifilter = ost->filter->graph->inputs[i];
if (!ifilter->ist->got_output && !input_files[ifilter->ist->file_index]->eof_reached) {
ist = ifilter->ist;
break;
}
}
if (!ist) {
ost->inputs_done = 1;
return 0;
}
} else {
ist = ost->ist;
av_assert0(ist);
}
ret = process_input(ist->file_index);
if (ret == AVERROR(EAGAIN)) {
if (input_files[ist->file_index]->eagain)
ost->unavailable = 1;
return 0;
}
if (ret < 0)
return ret == AVERROR_EOF ? 0 : ret;
return reap_filters(0);
}
/*
* The following code is the main loop of the file converter
*/
static int transcode(void)
{
int ret, i;
InputStream *ist;
int64_t timer_start;
ret = transcode_init();
if (ret < 0)
return ret;
if (stdin_interaction) {
av_log(NULL, AV_LOG_INFO, "Press [q] to stop, [?] for help\n");
}
timer_start = av_gettime_relative();
while (!received_sigterm) {
int64_t cur_time= av_gettime_relative();
/* if 'q' pressed, exits */
if (stdin_interaction)
if (check_keyboard_interaction(cur_time) < 0)
break;
/* check if there's any stream where output is still needed */
if (!need_output()) {
av_log(NULL, AV_LOG_VERBOSE, "No more output streams to write to, finishing.\n");
break;
}
ret = transcode_step();
if (ret < 0 && ret != AVERROR_EOF) {
av_log(NULL, AV_LOG_ERROR, "Error while filtering: %s\n", av_err2str(ret));
break;
}
/* dump report by using the output first video and audio streams */
print_report(0, timer_start, cur_time);
}
/* at the end of stream, we must flush the decoder buffers */
for (ist = ist_iter(NULL); ist; ist = ist_iter(ist)) {
if (!input_files[ist->file_index]->eof_reached) {
process_input_packet(ist, NULL, 0);
}
}
enc_flush();
term_exit();
/* write the trailer if needed */
for (i = 0; i < nb_output_files; i++) {
int err = of_write_trailer(output_files[i]);
ret = err_merge(ret, err);
}
/* dump report by using the first video and audio streams */
print_report(1, timer_start, av_gettime_relative());
return ret;
}
static BenchmarkTimeStamps get_benchmark_time_stamps(void)
{
BenchmarkTimeStamps time_stamps = { av_gettime_relative() };
#if HAVE_GETRUSAGE
struct rusage rusage;
2011-11-02 23:50:54 +03:00
getrusage(RUSAGE_SELF, &rusage);
time_stamps.user_usec =
(rusage.ru_utime.tv_sec * 1000000LL) + rusage.ru_utime.tv_usec;
time_stamps.sys_usec =
(rusage.ru_stime.tv_sec * 1000000LL) + rusage.ru_stime.tv_usec;
#elif HAVE_GETPROCESSTIMES
HANDLE proc;
FILETIME c, e, k, u;
proc = GetCurrentProcess();
GetProcessTimes(proc, &c, &e, &k, &u);
time_stamps.user_usec =
((int64_t)u.dwHighDateTime << 32 | u.dwLowDateTime) / 10;
time_stamps.sys_usec =
((int64_t)k.dwHighDateTime << 32 | k.dwLowDateTime) / 10;
#else
time_stamps.user_usec = time_stamps.sys_usec = 0;
#endif
return time_stamps;
}
static int64_t getmaxrss(void)
{
#if HAVE_GETRUSAGE && HAVE_STRUCT_RUSAGE_RU_MAXRSS
struct rusage rusage;
getrusage(RUSAGE_SELF, &rusage);
return (int64_t)rusage.ru_maxrss * 1024;
#elif HAVE_GETPROCESSMEMORYINFO
HANDLE proc;
PROCESS_MEMORY_COUNTERS memcounters;
proc = GetCurrentProcess();
memcounters.cb = sizeof(memcounters);
GetProcessMemoryInfo(proc, &memcounters, sizeof(memcounters));
return memcounters.PeakPagefileUsage;
#else
return 0;
#endif
}
int main(int argc, char **argv)
{
int ret;
BenchmarkTimeStamps ti;
init_dynload();
register_exit(ffmpeg_cleanup);
setvbuf(stderr,NULL,_IONBF,0); /* win32 runtime needs this */
av_log_set_flags(AV_LOG_SKIP_REPEATED);
parse_loglevel(argc, argv, options);
#if CONFIG_AVDEVICE
avdevice_register_all();
#endif
avformat_network_init();
show_banner(argc, argv, options);
2012-06-08 22:35:16 +03:00
/* parse options and open all input/output files */
ret = ffmpeg_parse_options(argc, argv);
2012-06-08 22:35:16 +03:00
if (ret < 0)
exit_program(1);
if (nb_output_files <= 0 && nb_input_files == 0) {
show_usage();
av_log(NULL, AV_LOG_WARNING, "Use -h to get full help or, even better, run 'man %s'\n", program_name);
exit_program(1);
}
/* file converter / grab */
if (nb_output_files <= 0) {
av_log(NULL, AV_LOG_FATAL, "At least one output file must be specified\n");
exit_program(1);
}
current_time = ti = get_benchmark_time_stamps();
ret = transcode();
if (ret >= 0 && do_benchmark) {
int64_t utime, stime, rtime;
current_time = get_benchmark_time_stamps();
utime = current_time.user_usec - ti.user_usec;
stime = current_time.sys_usec - ti.sys_usec;
rtime = current_time.real_usec - ti.real_usec;
av_log(NULL, AV_LOG_INFO,
"bench: utime=%0.3fs stime=%0.3fs rtime=%0.3fs\n",
utime / 1000000.0, stime / 1000000.0, rtime / 1000000.0);
}
av_log(NULL, AV_LOG_DEBUG, "%"PRIu64" frames successfully decoded, %"PRIu64" decoding errors\n",
decode_error_stat[0], decode_error_stat[1]);
if ((decode_error_stat[0] + decode_error_stat[1]) * max_error_rate < decode_error_stat[1])
exit_program(69);
ret = received_nb_signals ? 255 : ret;
exit_program(ret);
return ret;
}