mpd/src/decoder_thread.c

372 lines
8.3 KiB
C
Raw Normal View History

/*
* Copyright (C) 2003-2009 The Music Player Daemon Project
* http://www.musicpd.org
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program 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 General Public License for more details.
*
* You should have received a copy of the GNU General Public License along
* with this program; if not, write to the Free Software Foundation, Inc.,
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
*/
#include "decoder_thread.h"
2008-08-26 08:44:19 +02:00
#include "decoder_control.h"
#include "decoder_internal.h"
#include "decoder_list.h"
#include "decoder_plugin.h"
#include "input_stream.h"
#include "player_control.h"
#include "pipe.h"
#include "song.h"
#include "tag.h"
#include "mapper.h"
#include "path.h"
#include "uri.h"
2008-11-25 15:12:00 +01:00
#include <glib.h>
#include <unistd.h>
static bool
decoder_stream_decode(const struct decoder_plugin *plugin,
struct decoder *decoder,
struct input_stream *input_stream)
{
assert(plugin != NULL);
assert(plugin->stream_decode != NULL);
assert(decoder != NULL);
assert(decoder->stream_tag == NULL);
assert(decoder->decoder_tag == NULL);
assert(input_stream != NULL);
assert(input_stream->ready);
assert(decoder->dc->state == DECODE_STATE_START);
decoder_unlock(decoder->dc);
/* rewind the stream, so each plugin gets a fresh start */
input_stream_seek(input_stream, 0, SEEK_SET);
decoder_plugin_stream_decode(plugin, decoder, input_stream);
decoder_lock(decoder->dc);
assert(decoder->dc->state == DECODE_STATE_START ||
decoder->dc->state == DECODE_STATE_DECODE);
return decoder->dc->state != DECODE_STATE_START;
}
static bool
decoder_file_decode(const struct decoder_plugin *plugin,
struct decoder *decoder, const char *path)
{
assert(plugin != NULL);
assert(plugin->file_decode != NULL);
assert(decoder != NULL);
assert(decoder->stream_tag == NULL);
assert(decoder->decoder_tag == NULL);
assert(path != NULL);
assert(g_path_is_absolute(path));
assert(decoder->dc->state == DECODE_STATE_START);
decoder_unlock(decoder->dc);
decoder_plugin_file_decode(plugin, decoder, path);
decoder_lock(decoder->dc);
assert(decoder->dc->state == DECODE_STATE_START ||
decoder->dc->state == DECODE_STATE_DECODE);
return decoder->dc->state != DECODE_STATE_START;
}
/**
* Try decoding a stream, using plugins matching the stream's MIME type.
*/
static bool
decoder_run_stream_mime_type(struct decoder *decoder, struct input_stream *is)
{
const struct decoder_plugin *plugin;
unsigned int next = 0;
if (is->mime == NULL)
return false;
while ((plugin = decoder_plugin_from_mime_type(is->mime, next++)))
if (plugin->stream_decode != NULL &&
decoder_stream_decode(plugin, decoder, is))
return true;
return false;
}
/**
* Try decoding a stream, using plugins matching the stream's URI
* suffix.
*/
static bool
decoder_run_stream_suffix(struct decoder *decoder, struct input_stream *is,
const char *uri)
{
const char *suffix = uri_get_suffix(uri);
const struct decoder_plugin *plugin = NULL;
if (suffix == NULL)
return false;
while ((plugin = decoder_plugin_from_suffix(suffix, plugin)) != NULL)
if (plugin->stream_decode != NULL &&
decoder_stream_decode(plugin, decoder, is))
return true;
return false;
}
/**
* Try decoding a stream, using the fallback plugin.
*/
static bool
decoder_run_stream_fallback(struct decoder *decoder, struct input_stream *is)
{
const struct decoder_plugin *plugin;
plugin = decoder_plugin_from_name("mad");
return plugin != NULL && plugin->stream_decode != NULL &&
decoder_stream_decode(plugin, decoder, is);
}
/**
* Try decoding a stream.
*/
static bool
decoder_run_stream(struct decoder *decoder, const char *uri)
{
struct decoder_control *dc = decoder->dc;
struct input_stream input_stream;
bool success;
decoder_unlock(dc);
if (!input_stream_open(&input_stream, uri)) {
decoder_lock(dc);
dc->state = DECODE_STATE_ERROR;
return false;
}
/* wait for the input stream to become ready; its metadata
will be available then */
while (!input_stream.ready) {
int ret;
decoder_lock(dc);
if (dc->command == DECODE_COMMAND_STOP) {
decoder_unlock(dc);
input_stream_close(&input_stream);
decoder_lock(dc);
return true;
}
decoder_unlock(dc);
ret = input_stream_buffer(&input_stream);
if (ret < 0) {
input_stream_close(&input_stream);
decoder_lock(dc);
return false;
}
}
decoder_lock(dc);
success = dc->command == DECODE_COMMAND_STOP ||
/* first we try mime types: */
decoder_run_stream_mime_type(decoder, &input_stream) ||
/* if that fails, try suffix matching the URL: */
decoder_run_stream_suffix(decoder, &input_stream, uri) ||
/* fallback to mp3: this is needed for bastard streams
that don't have a suffix or set the mimeType */
decoder_run_stream_fallback(decoder, &input_stream);
decoder_unlock(dc);
input_stream_close(&input_stream);
decoder_lock(dc);
return success;
}
/**
* Try decoding a file.
*/
static bool
decoder_run_file(struct decoder *decoder, const char *path_fs)
{
const char *suffix = uri_get_suffix(path_fs);
const struct decoder_plugin *plugin = NULL;
if (suffix == NULL)
return false;
decoder_unlock(decoder->dc);
while ((plugin = decoder_plugin_from_suffix(suffix, plugin)) != NULL) {
if (plugin->file_decode != NULL) {
decoder_lock(decoder->dc);
if (decoder_file_decode(plugin, decoder, path_fs))
return true;
decoder_unlock(decoder->dc);
} else if (plugin->stream_decode != NULL) {
struct input_stream input_stream;
if (!input_stream_open(&input_stream, path_fs))
continue;
decoder_lock(decoder->dc);
if (decoder_stream_decode(plugin, decoder,
&input_stream))
return true;
decoder_unlock(decoder->dc);
}
}
decoder_lock(decoder->dc);
return false;
}
static void
decoder_run_song(struct decoder_control *dc,
const struct song *song, const char *uri)
{
struct decoder decoder = {
.dc = dc,
};
int ret;
decoder.seeking = false;
decoder.song_tag = song->tag != NULL && song_is_file(song)
? tag_dup(song->tag) : NULL;
decoder.stream_tag = NULL;
decoder.decoder_tag = NULL;
decoder.chunk = NULL;
dc->state = DECODE_STATE_START;
dc->command = DECODE_COMMAND_NONE;
player_signal();
pcm_convert_init(&decoder.conv_state);
ret = song_is_file(song)
? decoder_run_file(&decoder, uri)
: decoder_run_stream(&decoder, uri);
decoder_unlock(dc);
pcm_convert_deinit(&decoder.conv_state);
/* flush the last chunk */
if (decoder.chunk != NULL)
decoder_flush_chunk(&decoder);
if (decoder.song_tag != NULL)
tag_free(decoder.song_tag);
if (decoder.stream_tag != NULL)
tag_free(decoder.stream_tag);
if (decoder.decoder_tag != NULL)
tag_free(decoder.decoder_tag);
decoder_lock(dc);
dc->state = ret ? DECODE_STATE_STOP : DECODE_STATE_ERROR;
}
static void
decoder_run(struct decoder_control *dc)
{
const struct song *song = dc->song;
char *uri;
assert(song != NULL);
if (song_is_file(song))
uri = map_song_fs(song);
else
uri = song_get_uri(song);
if (uri == NULL) {
dc->state = DECODE_STATE_ERROR;
return;
}
decoder_run_song(dc, song, uri);
g_free(uri);
}
static gpointer
decoder_task(gpointer arg)
{
struct decoder_control *dc = arg;
decoder_lock(dc);
do {
assert(dc->state == DECODE_STATE_STOP ||
dc->state == DECODE_STATE_ERROR);
switch (dc->command) {
case DECODE_COMMAND_START:
case DECODE_COMMAND_SEEK:
decoder_run(dc);
dc->command = DECODE_COMMAND_NONE;
player_signal();
break;
case DECODE_COMMAND_STOP:
dc->command = DECODE_COMMAND_NONE;
player_signal();
break;
case DECODE_COMMAND_NONE:
decoder_wait(dc);
break;
}
} while (dc->command != DECODE_COMMAND_NONE || !dc->quit);
decoder_unlock(dc);
return NULL;
Initial cut of fork() => pthreads() for decoder and player I initially started to do a heavy rewrite that changed the way processes communicated, but that was too much to do at once. So this change only focuses on replacing the player and decode processes with threads and using condition variables instead of polling in loops; so the changeset itself is quiet small. * The shared output buffer variables will still need locking to guard against race conditions. So in this effect, we're probably just as buggy as before. The reduced context-switching overhead of using threads instead of processes may even make bugs show up more or less often... * Basic functionality appears to be working for playing local (and NFS) audio, including: play, pause, stop, seek, previous, next, and main playlist editing * I haven't tested HTTP streams yet, they should work. * I've only tested ALSA and Icecast. ALSA works fine, Icecast metadata seems to get screwy at times and breaks song advancement in the playlist at times. * state file loading works, too (after some last-minute hacks with non-blocking wakeup functions) * The non-blocking (*_nb) variants of the task management functions are probably overused. They're more lenient and easier to use because much of our code is still based on our previous polling-based system. * It currently segfaults on exit. I haven't paid much attention to the exit/signal-handling routines other than ensuring it compiles. At least the state file seems to work. We don't do any cleanups of the threads on exit, yet. * Update is still done in a child process and not in a thread. To do this in a thread, we'll need to ensure it does proper locking and communication with the main thread; but should require less memory in the end because we'll be updating the database "in-place" rather than updating a copy and then bulk-loading when done. * We're more sensitive to bugs in 3rd party libraries now. My plan is to eventually use a master process which forks() and restarts the child when it dies: locking and communication with the main thread; but should require less memory in the end because we'll be updating the database "in-place" rather than updating a copy and then bulk-loading when done. * We're more sensitive to bugs in 3rd party libraries now. My plan is to eventually use a master process which forks() and restarts the child when it dies: master - just does waitpid() + fork() in a loop \- main thread \- decoder thread \- player thread At the beginning of every song, the main thread will set a dirty flag and update the state file. This way, if we encounter a song that triggers a segfault killing the main thread, the master will start the replacement main on the next song. * The main thread still wakes up every second on select() to check for signals; which affects power management. [merged r7138 from branches/ew] git-svn-id: https://svn.musicpd.org/mpd/trunk@7240 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12 06:08:00 +02:00
}
void
decoder_thread_start(struct decoder_control *dc)
Initial cut of fork() => pthreads() for decoder and player I initially started to do a heavy rewrite that changed the way processes communicated, but that was too much to do at once. So this change only focuses on replacing the player and decode processes with threads and using condition variables instead of polling in loops; so the changeset itself is quiet small. * The shared output buffer variables will still need locking to guard against race conditions. So in this effect, we're probably just as buggy as before. The reduced context-switching overhead of using threads instead of processes may even make bugs show up more or less often... * Basic functionality appears to be working for playing local (and NFS) audio, including: play, pause, stop, seek, previous, next, and main playlist editing * I haven't tested HTTP streams yet, they should work. * I've only tested ALSA and Icecast. ALSA works fine, Icecast metadata seems to get screwy at times and breaks song advancement in the playlist at times. * state file loading works, too (after some last-minute hacks with non-blocking wakeup functions) * The non-blocking (*_nb) variants of the task management functions are probably overused. They're more lenient and easier to use because much of our code is still based on our previous polling-based system. * It currently segfaults on exit. I haven't paid much attention to the exit/signal-handling routines other than ensuring it compiles. At least the state file seems to work. We don't do any cleanups of the threads on exit, yet. * Update is still done in a child process and not in a thread. To do this in a thread, we'll need to ensure it does proper locking and communication with the main thread; but should require less memory in the end because we'll be updating the database "in-place" rather than updating a copy and then bulk-loading when done. * We're more sensitive to bugs in 3rd party libraries now. My plan is to eventually use a master process which forks() and restarts the child when it dies: locking and communication with the main thread; but should require less memory in the end because we'll be updating the database "in-place" rather than updating a copy and then bulk-loading when done. * We're more sensitive to bugs in 3rd party libraries now. My plan is to eventually use a master process which forks() and restarts the child when it dies: master - just does waitpid() + fork() in a loop \- main thread \- decoder thread \- player thread At the beginning of every song, the main thread will set a dirty flag and update the state file. This way, if we encounter a song that triggers a segfault killing the main thread, the master will start the replacement main on the next song. * The main thread still wakes up every second on select() to check for signals; which affects power management. [merged r7138 from branches/ew] git-svn-id: https://svn.musicpd.org/mpd/trunk@7240 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12 06:08:00 +02:00
{
GError *e = NULL;
Initial cut of fork() => pthreads() for decoder and player I initially started to do a heavy rewrite that changed the way processes communicated, but that was too much to do at once. So this change only focuses on replacing the player and decode processes with threads and using condition variables instead of polling in loops; so the changeset itself is quiet small. * The shared output buffer variables will still need locking to guard against race conditions. So in this effect, we're probably just as buggy as before. The reduced context-switching overhead of using threads instead of processes may even make bugs show up more or less often... * Basic functionality appears to be working for playing local (and NFS) audio, including: play, pause, stop, seek, previous, next, and main playlist editing * I haven't tested HTTP streams yet, they should work. * I've only tested ALSA and Icecast. ALSA works fine, Icecast metadata seems to get screwy at times and breaks song advancement in the playlist at times. * state file loading works, too (after some last-minute hacks with non-blocking wakeup functions) * The non-blocking (*_nb) variants of the task management functions are probably overused. They're more lenient and easier to use because much of our code is still based on our previous polling-based system. * It currently segfaults on exit. I haven't paid much attention to the exit/signal-handling routines other than ensuring it compiles. At least the state file seems to work. We don't do any cleanups of the threads on exit, yet. * Update is still done in a child process and not in a thread. To do this in a thread, we'll need to ensure it does proper locking and communication with the main thread; but should require less memory in the end because we'll be updating the database "in-place" rather than updating a copy and then bulk-loading when done. * We're more sensitive to bugs in 3rd party libraries now. My plan is to eventually use a master process which forks() and restarts the child when it dies: locking and communication with the main thread; but should require less memory in the end because we'll be updating the database "in-place" rather than updating a copy and then bulk-loading when done. * We're more sensitive to bugs in 3rd party libraries now. My plan is to eventually use a master process which forks() and restarts the child when it dies: master - just does waitpid() + fork() in a loop \- main thread \- decoder thread \- player thread At the beginning of every song, the main thread will set a dirty flag and update the state file. This way, if we encounter a song that triggers a segfault killing the main thread, the master will start the replacement main on the next song. * The main thread still wakes up every second on select() to check for signals; which affects power management. [merged r7138 from branches/ew] git-svn-id: https://svn.musicpd.org/mpd/trunk@7240 09075e82-0dd4-0310-85a5-a0d7c8717e4f
2008-04-12 06:08:00 +02:00
assert(dc->thread == NULL);
dc->quit = false;
dc->thread = g_thread_create(decoder_task, dc, true, &e);
if (dc->thread == NULL)
g_error("Failed to spawn decoder task: %s", e->message);
}