Add three parameters to the backend play() function call -- (1) a flag indicating whether the samples are timed or not. If timed, (2) the timestamp and (3) the local time at which the first frame should be heard.
Update missing pipewire library message.
Makefile.am fix.
Update the SHM version and fix the SHM name so that it works with FreeBSD.
Remove redundant (?) AC_HEADER_STDC check.
Fix compilation and installation under FreeBSD -- changes to allow building in a separate directory broke the FreeBSD build process.
Add AC_CHECK_LIB for gcrypt in case PKG_CHECK_MODULES fails to find it.
Find gcrypt using pkg-config
Add code to detect when the RTSP channel goes idle for a period, and use it to reset clocks and timings, etc. so that SPS resumes correctly where the source has gone to sleep while playing a realtime stream and has subsequently woken up.
Don't exit if a UDP Clock Control packet is empty. Also check minimum packet size.
Remove some very experimental code which may cause memory and port leaks.
Strip file path from the filename used in debug messages, information messages, warnings and fatal error messages.
BB fix to Makefile.am -- overwriting the DBUS stuff with MPRIS definitions, duh.
Fix some problems building on FreeBSD and tidy up the use of the "sed" editor.
Fix some problems building on picore.
Fix an uninitialised variable
start using the "nqptp" SMI
451 lines
16 KiB
C
451 lines
16 KiB
C
/*
|
|
* Asynchronous PulseAudio Backend. This file is part of Shairport Sync.
|
|
* Copyright (c) Mike Brady 2017
|
|
* All rights reserved.
|
|
*
|
|
* Permission is hereby granted, free of charge, to any person
|
|
* obtaining a copy of this software and associated documentation
|
|
* files (the "Software"), to deal in the Software without
|
|
* restriction, including without limitation the rights to use,
|
|
* copy, modify, merge, publish, distribute, sublicense, and/or
|
|
* sell copies of the Software, and to permit persons to whom the
|
|
* Software is furnished to do so, subject to the following conditions:
|
|
*
|
|
* The above copyright notice and this permission notice shall be
|
|
* included in all copies or substantial portions of the Software.
|
|
*
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
|
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
|
|
* OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
|
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
|
|
* HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
|
|
* WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
|
|
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
|
|
* OTHER DEALINGS IN THE SOFTWARE.
|
|
*/
|
|
|
|
// Based (distantly, with thanks) on
|
|
// http://stackoverflow.com/questions/29977651/how-can-the-pulseaudio-asynchronous-library-be-used-to-play-raw-pcm-data
|
|
|
|
#include "audio.h"
|
|
#include "common.h"
|
|
#include <errno.h>
|
|
#include <pthread.h>
|
|
#include <pulse/pulseaudio.h>
|
|
#include <stdio.h>
|
|
#include <string.h>
|
|
#include <unistd.h>
|
|
|
|
// note -- these are hacked and hardwired into this code.
|
|
#define FORMAT PA_SAMPLE_S16NE
|
|
#define RATE 44100
|
|
|
|
// Four seconds buffer -- should be plenty
|
|
#define buffer_allocation 44100 * 4 * 2 * 2
|
|
|
|
static pthread_mutex_t buffer_mutex = PTHREAD_MUTEX_INITIALIZER;
|
|
|
|
/*
|
|
static struct {
|
|
char *server;
|
|
char *sink;
|
|
char *service_name;
|
|
} pulse_options = {.server = NULL, .sink = NULL, .service_name = NULL};
|
|
*/
|
|
|
|
pa_threaded_mainloop *mainloop;
|
|
pa_mainloop_api *mainloop_api;
|
|
pa_context *context;
|
|
pa_stream *stream;
|
|
char *audio_lmb, *audio_umb, *audio_toq, *audio_eoq;
|
|
size_t audio_size = buffer_allocation;
|
|
size_t audio_occupancy;
|
|
|
|
void context_state_cb(pa_context *context, void *mainloop);
|
|
void stream_state_cb(pa_stream *s, void *mainloop);
|
|
void stream_success_cb(pa_stream *stream, int success, void *userdata);
|
|
void stream_write_cb(pa_stream *stream, size_t requested_bytes, void *userdata);
|
|
|
|
static int init(__attribute__((unused)) int argc, __attribute__((unused)) char **argv) {
|
|
|
|
// set up default values first
|
|
config.audio_backend_buffer_desired_length = 0.35;
|
|
config.audio_backend_buffer_interpolation_threshold_in_seconds =
|
|
0.02; // below this, soxr interpolation will not occur -- it'll be basic interpolation
|
|
// instead.
|
|
|
|
config.audio_backend_latency_offset = 0;
|
|
|
|
// get settings from settings file
|
|
|
|
// do the "general" audio options. Note, these options are in the "general" stanza!
|
|
parse_general_audio_options();
|
|
|
|
// now the specific options
|
|
if (config.cfg != NULL) {
|
|
const char *str;
|
|
|
|
/* Get the PulseAudio server name. */
|
|
if (config_lookup_string(config.cfg, "pa.server", &str)) {
|
|
config.pa_server = (char *)str;
|
|
}
|
|
|
|
/* Get the Application Name. */
|
|
if (config_lookup_string(config.cfg, "pa.application_name", &str)) {
|
|
config.pa_application_name = (char *)str;
|
|
}
|
|
|
|
/* Get the PulseAudio sink name. */
|
|
if (config_lookup_string(config.cfg, "pa.sink", &str)) {
|
|
config.pa_sink = (char *)str;
|
|
}
|
|
}
|
|
|
|
// finish collecting settings
|
|
|
|
// allocate space for the audio buffer
|
|
audio_lmb = malloc(audio_size);
|
|
if (audio_lmb == NULL)
|
|
die("Can't allocate %d bytes for pulseaudio buffer.", audio_size);
|
|
audio_toq = audio_eoq = audio_lmb;
|
|
audio_umb = audio_lmb + audio_size;
|
|
audio_occupancy = 0;
|
|
|
|
// Get a mainloop and its context
|
|
mainloop = pa_threaded_mainloop_new();
|
|
if (mainloop == NULL)
|
|
die("could not create a pa_threaded_mainloop.");
|
|
mainloop_api = pa_threaded_mainloop_get_api(mainloop);
|
|
if (config.pa_application_name)
|
|
context = pa_context_new(mainloop_api, config.pa_application_name);
|
|
else
|
|
context = pa_context_new(mainloop_api, "Shairport Sync");
|
|
if (context == NULL)
|
|
die("could not create a new context for pulseaudio.");
|
|
// Set a callback so we can wait for the context to be ready
|
|
pa_context_set_state_callback(context, &context_state_cb, mainloop);
|
|
|
|
// Lock the mainloop so that it does not run and crash before the context is ready
|
|
pa_threaded_mainloop_lock(mainloop);
|
|
|
|
// Start the mainloop
|
|
if (pa_threaded_mainloop_start(mainloop) != 0)
|
|
die("could not start the pulseaudio threaded mainloop");
|
|
|
|
if (pa_context_connect(context, config.pa_server, 0, NULL) != 0)
|
|
die("failed to connect to the pulseaudio context -- the error message is \"%s\".",
|
|
pa_strerror(pa_context_errno(context)));
|
|
|
|
// Wait for the context to be ready
|
|
for (;;) {
|
|
pa_context_state_t context_state = pa_context_get_state(context);
|
|
if (!PA_CONTEXT_IS_GOOD(context_state))
|
|
die("pa context is not good -- the error message \"%s\".",
|
|
pa_strerror(pa_context_errno(context)));
|
|
if (context_state == PA_CONTEXT_READY)
|
|
break;
|
|
pa_threaded_mainloop_wait(mainloop);
|
|
}
|
|
|
|
pa_threaded_mainloop_unlock(mainloop);
|
|
|
|
return 0;
|
|
}
|
|
|
|
static void deinit(void) {
|
|
pa_threaded_mainloop_stop(mainloop);
|
|
pa_threaded_mainloop_free(mainloop);
|
|
// debug(1, "pa deinit done");
|
|
}
|
|
|
|
static void start(__attribute__((unused)) int sample_rate,
|
|
__attribute__((unused)) int sample_format) {
|
|
|
|
uint32_t buffer_size_in_bytes = (uint32_t)2 * 2 * RATE * 0.1; // hard wired in here
|
|
// debug(1, "pa_buffer size is %u bytes.", buffer_size_in_bytes);
|
|
|
|
pa_threaded_mainloop_lock(mainloop);
|
|
// Create a playback stream
|
|
pa_sample_spec sample_specifications;
|
|
sample_specifications.format = FORMAT;
|
|
sample_specifications.rate = RATE;
|
|
sample_specifications.channels = 2;
|
|
|
|
pa_channel_map map;
|
|
pa_channel_map_init_stereo(&map);
|
|
|
|
stream = pa_stream_new(context, "Playback", &sample_specifications, &map);
|
|
pa_stream_set_state_callback(stream, stream_state_cb, mainloop);
|
|
pa_stream_set_write_callback(stream, stream_write_cb, mainloop);
|
|
// pa_stream_set_latency_update_callback(stream, stream_latency_cb, mainloop);
|
|
|
|
// recommended settings, i.e. server uses sensible values
|
|
pa_buffer_attr buffer_attr;
|
|
buffer_attr.maxlength = (uint32_t)-1;
|
|
buffer_attr.tlength = buffer_size_in_bytes;
|
|
buffer_attr.prebuf = (uint32_t)0;
|
|
buffer_attr.minreq = (uint32_t)-1;
|
|
|
|
// Settings copied as per the chromium browser source
|
|
pa_stream_flags_t stream_flags;
|
|
stream_flags = PA_STREAM_START_CORKED | PA_STREAM_INTERPOLATE_TIMING | PA_STREAM_NOT_MONOTONIC |
|
|
// PA_STREAM_AUTO_TIMING_UPDATE;
|
|
PA_STREAM_AUTO_TIMING_UPDATE | PA_STREAM_ADJUST_LATENCY;
|
|
|
|
int connect_result;
|
|
|
|
if (config.pa_sink) {
|
|
// Connect stream to the sink specified in the config
|
|
connect_result =
|
|
pa_stream_connect_playback(stream, config.pa_sink, &buffer_attr, stream_flags, NULL, NULL);
|
|
} else {
|
|
// Connect stream to the default audio output sink
|
|
connect_result =
|
|
pa_stream_connect_playback(stream, NULL, &buffer_attr, stream_flags, NULL, NULL);
|
|
}
|
|
|
|
if (connect_result != 0)
|
|
die("could not connect to the pulseaudio playback stream -- the error message is \"%s\".",
|
|
pa_strerror(pa_context_errno(context)));
|
|
|
|
// Wait for the stream to be ready
|
|
for (;;) {
|
|
pa_stream_state_t stream_state = pa_stream_get_state(stream);
|
|
if (!PA_STREAM_IS_GOOD(stream_state))
|
|
die("stream state is no longer good while waiting for stream to become ready -- the error "
|
|
"message is \"%s\".",
|
|
pa_strerror(pa_context_errno(context)));
|
|
if (stream_state == PA_STREAM_READY)
|
|
break;
|
|
pa_threaded_mainloop_wait(mainloop);
|
|
}
|
|
|
|
pa_threaded_mainloop_unlock(mainloop);
|
|
}
|
|
|
|
static int play(void *buf, int samples, __attribute__((unused)) int sample_type,
|
|
__attribute__((unused)) uint32_t timestamp,
|
|
__attribute__((unused)) uint64_t playtime) {
|
|
// debug(1,"pa_play of %d samples.",samples);
|
|
// copy the samples into the queue
|
|
size_t bytes_to_transfer = samples * 2 * 2;
|
|
size_t space_to_end_of_buffer = audio_umb - audio_eoq;
|
|
if (space_to_end_of_buffer >= bytes_to_transfer) {
|
|
memcpy(audio_eoq, buf, bytes_to_transfer);
|
|
audio_occupancy += bytes_to_transfer;
|
|
pthread_mutex_lock(&buffer_mutex);
|
|
audio_eoq += bytes_to_transfer;
|
|
pthread_mutex_unlock(&buffer_mutex);
|
|
} else {
|
|
memcpy(audio_eoq, buf, space_to_end_of_buffer);
|
|
buf += space_to_end_of_buffer;
|
|
memcpy(audio_lmb, buf, bytes_to_transfer - space_to_end_of_buffer);
|
|
pthread_mutex_lock(&buffer_mutex);
|
|
audio_occupancy += bytes_to_transfer;
|
|
pthread_mutex_unlock(&buffer_mutex);
|
|
audio_eoq = audio_lmb + bytes_to_transfer - space_to_end_of_buffer;
|
|
}
|
|
if ((audio_occupancy >= 11025 * 2 * 2) && (pa_stream_is_corked(stream))) {
|
|
// debug(1,"Uncorked");
|
|
pa_threaded_mainloop_lock(mainloop);
|
|
pa_stream_cork(stream, 0, stream_success_cb, mainloop);
|
|
pa_threaded_mainloop_unlock(mainloop);
|
|
}
|
|
return 0;
|
|
}
|
|
|
|
int pa_delay(long *the_delay) {
|
|
long result = 0;
|
|
int reply = -ENODEV;
|
|
pa_usec_t latency;
|
|
int negative;
|
|
pa_threaded_mainloop_lock(mainloop);
|
|
int gl = pa_stream_get_latency(stream, &latency, &negative);
|
|
pa_threaded_mainloop_unlock(mainloop);
|
|
if (gl == PA_ERR_NODATA) {
|
|
// debug(1, "No latency data yet.");
|
|
reply = -ENODEV;
|
|
} else if (gl != 0) {
|
|
// debug(1,"Error %d getting latency.",gl);
|
|
reply = -EIO;
|
|
} else {
|
|
result = (audio_occupancy / (2 * 2)) + (latency * 44100) / 1000000;
|
|
reply = 0;
|
|
}
|
|
*the_delay = result;
|
|
return reply;
|
|
}
|
|
|
|
void flush(void) {
|
|
// Cork the stream so it will stop playing
|
|
pa_threaded_mainloop_lock(mainloop);
|
|
if (pa_stream_is_corked(stream) == 0) {
|
|
// debug(1,"Flush and cork for flush.");
|
|
pa_stream_flush(stream, stream_success_cb, NULL);
|
|
pa_stream_cork(stream, 1, stream_success_cb, mainloop);
|
|
}
|
|
pa_threaded_mainloop_unlock(mainloop);
|
|
audio_toq = audio_eoq = audio_lmb;
|
|
audio_umb = audio_lmb + audio_size;
|
|
audio_occupancy = 0;
|
|
}
|
|
|
|
static void stop(void) {
|
|
// Cork the stream so it will stop playing
|
|
pa_threaded_mainloop_lock(mainloop);
|
|
if (pa_stream_is_corked(stream) == 0) {
|
|
// debug(1,"Flush and cork for stop.");
|
|
pa_stream_flush(stream, stream_success_cb, NULL);
|
|
pa_stream_cork(stream, 1, stream_success_cb, mainloop);
|
|
}
|
|
pa_threaded_mainloop_unlock(mainloop);
|
|
audio_toq = audio_eoq = audio_lmb;
|
|
audio_umb = audio_lmb + audio_size;
|
|
audio_occupancy = 0;
|
|
|
|
// debug(1,"pa stop");
|
|
pa_stream_disconnect(stream);
|
|
}
|
|
|
|
audio_output audio_pa = {.name = "pa",
|
|
.help = NULL,
|
|
.init = &init,
|
|
.deinit = &deinit,
|
|
.prepare = NULL,
|
|
.start = &start,
|
|
.stop = &stop,
|
|
.is_running = NULL,
|
|
.flush = &flush,
|
|
.delay = &pa_delay,
|
|
.stats = NULL,
|
|
.play = &play,
|
|
.volume = NULL,
|
|
.parameters = NULL,
|
|
.mute = NULL};
|
|
|
|
void context_state_cb(__attribute__((unused)) pa_context *context, void *mainloop) {
|
|
pa_threaded_mainloop_signal(mainloop, 0);
|
|
}
|
|
|
|
void stream_state_cb(__attribute__((unused)) pa_stream *s, void *mainloop) {
|
|
pa_threaded_mainloop_signal(mainloop, 0);
|
|
}
|
|
|
|
void stream_write_cb(pa_stream *stream, size_t requested_bytes,
|
|
__attribute__((unused)) void *userdata) {
|
|
|
|
/*
|
|
// play with timing information
|
|
const struct pa_timing_info *ti = pa_stream_get_timing_info(stream);
|
|
if ((ti == NULL) || (ti->write_index_corrupt)) {
|
|
debug(2, "Timing info invalid");
|
|
} else {
|
|
struct timeval time_now;
|
|
|
|
pa_gettimeofday(&time_now);
|
|
|
|
uint64_t time_now_fp = ((uint64_t)time_now.tv_sec << 32) +
|
|
((uint64_t)time_now.tv_usec << 32) / 1000000; // types okay
|
|
uint64_t time_of_ti_fp = ((uint64_t)(ti->timestamp.tv_sec) << 32) +
|
|
((uint64_t)(ti->timestamp.tv_usec) << 32) / 1000000; // types okay
|
|
|
|
if (time_now_fp >= time_of_ti_fp) {
|
|
uint64_t estimate_age = ((time_now_fp - time_of_ti_fp) * 1000000) >> 32;
|
|
uint64_t bytes_in_buffer = ti->write_index - ti->read_index;
|
|
pa_usec_t microseconds_to_write_buffer = (bytes_in_buffer * 1000000) / (44100 * 2 * 2);
|
|
pa_usec_t ea = (pa_usec_t)estimate_age;
|
|
pa_usec_t pa_latency = ti->sink_usec + ti->transport_usec + microseconds_to_write_buffer;
|
|
pa_usec_t estimated_latency = pa_latency - estimate_age;
|
|
// debug(1,"Estimated latency is %d microseconds.",estimated_latency);
|
|
|
|
// } else {
|
|
// debug(1, "Time now is earlier than time of timing information");
|
|
}
|
|
}
|
|
*/
|
|
int bytes_to_transfer = requested_bytes;
|
|
int bytes_transferred = 0;
|
|
uint8_t *buffer = NULL;
|
|
|
|
while ((bytes_to_transfer > 0) && (audio_occupancy > 0)) {
|
|
size_t bytes_we_can_transfer = bytes_to_transfer;
|
|
if (audio_occupancy < bytes_we_can_transfer) {
|
|
// debug(1, "Underflow? We have %d bytes but we are asked for %d bytes", audio_occupancy,
|
|
// bytes_we_can_transfer);
|
|
pa_stream_cork(stream, 1, stream_success_cb, mainloop);
|
|
// debug(1, "Corked");
|
|
bytes_we_can_transfer = audio_occupancy;
|
|
}
|
|
|
|
// bytes we can transfer will never be greater than the bytes available
|
|
|
|
pa_stream_begin_write(stream, (void **)&buffer, &bytes_we_can_transfer);
|
|
if (bytes_we_can_transfer <= (size_t)(audio_umb - audio_toq)) {
|
|
// the bytes are all in a row in the audo buffer
|
|
memcpy(buffer, audio_toq, bytes_we_can_transfer);
|
|
audio_toq += bytes_we_can_transfer;
|
|
// lock
|
|
pthread_mutex_lock(&buffer_mutex);
|
|
audio_occupancy -= bytes_we_can_transfer;
|
|
pthread_mutex_unlock(&buffer_mutex);
|
|
// unlock
|
|
pa_stream_write(stream, buffer, bytes_we_can_transfer, NULL, 0LL, PA_SEEK_RELATIVE);
|
|
bytes_transferred += bytes_we_can_transfer;
|
|
} else {
|
|
// the bytes are in two places in the audio buffer
|
|
size_t first_portion_to_write = audio_umb - audio_toq;
|
|
if (first_portion_to_write != 0)
|
|
memcpy(buffer, audio_toq, first_portion_to_write);
|
|
uint8_t *new_buffer = buffer + first_portion_to_write;
|
|
memcpy(new_buffer, audio_lmb, bytes_we_can_transfer - first_portion_to_write);
|
|
pa_stream_write(stream, buffer, bytes_we_can_transfer, NULL, 0LL, PA_SEEK_RELATIVE);
|
|
bytes_transferred += bytes_we_can_transfer;
|
|
audio_toq = audio_lmb + bytes_we_can_transfer - first_portion_to_write;
|
|
// lock
|
|
pthread_mutex_lock(&buffer_mutex);
|
|
audio_occupancy -= bytes_we_can_transfer;
|
|
pthread_mutex_unlock(&buffer_mutex);
|
|
// unlock
|
|
}
|
|
bytes_to_transfer -= bytes_we_can_transfer;
|
|
// debug(1,"audio_toq is %llx",audio_toq);
|
|
}
|
|
|
|
// debug(1,"<<<Frames requested %d, written to pa: %d, corked status:
|
|
// %d.",requested_bytes/4,bytes_transferred/4,pa_stream_is_corked(stream));
|
|
}
|
|
|
|
void alt_stream_write_cb(pa_stream *stream, size_t requested_bytes,
|
|
__attribute__((unused)) void *userdata) {
|
|
// debug(1, "***Bytes requested bytes %d.", requested_bytes);
|
|
size_t bytes_remaining = requested_bytes;
|
|
while (bytes_remaining > 0) {
|
|
uint8_t *buffer = NULL;
|
|
size_t bytes_to_fill = 44100;
|
|
size_t i;
|
|
|
|
if (bytes_to_fill > bytes_remaining)
|
|
bytes_to_fill = bytes_remaining;
|
|
|
|
pa_stream_begin_write(stream, (void **)&buffer, &bytes_to_fill);
|
|
if (buffer) {
|
|
for (i = 0; i < bytes_to_fill; i += 2) {
|
|
buffer[i] = (i % 100) * 40 / 100 + 44;
|
|
buffer[i + 1] = (i % 100) * 40 / 100 + 44;
|
|
}
|
|
} else {
|
|
die("buffer not allocated in alt_stream_write_cb.");
|
|
}
|
|
|
|
pa_stream_write(stream, buffer, bytes_to_fill, NULL, 0LL, PA_SEEK_RELATIVE);
|
|
|
|
bytes_remaining -= bytes_to_fill;
|
|
}
|
|
}
|
|
|
|
void stream_success_cb(__attribute__((unused)) pa_stream *stream,
|
|
__attribute__((unused)) int success,
|
|
__attribute__((unused)) void *userdata) {
|
|
return;
|
|
}
|