mirror of
https://gitlab.freedesktop.org/pulseaudio/pulseaudio.git
synced 2025-11-07 13:30:03 -05:00
FSF addresses used in PA sources are no longer valid and rpmlint generates numerous warnings during packaging because of this. This patch changes all FSF addresses to FSF web page according to the GPL how-to: https://www.gnu.org/licenses/gpl-howto.en.html Done automatically by sed-ing through sources.
218 lines
4.7 KiB
C
218 lines
4.7 KiB
C
/***
|
|
This file is part of PulseAudio.
|
|
|
|
Copyright 2006 Lennart Poettering
|
|
|
|
PulseAudio 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.
|
|
|
|
PulseAudio 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 PulseAudio; if not, see <http://www.gnu.org/licenses/>.
|
|
***/
|
|
|
|
#ifdef HAVE_CONFIG_H
|
|
#include <config.h>
|
|
#endif
|
|
|
|
#include <unistd.h>
|
|
#include <errno.h>
|
|
|
|
#include <pulsecore/atomic.h>
|
|
#include <pulsecore/log.h>
|
|
#include <pulsecore/thread.h>
|
|
#include <pulsecore/macro.h>
|
|
#include <pulsecore/core-util.h>
|
|
#include <pulse/xmalloc.h>
|
|
|
|
#include "fdsem.h"
|
|
|
|
/* For debugging purposes we can define _Y to put and extra thread
|
|
* yield between each operation. */
|
|
|
|
/* #define PROFILE */
|
|
|
|
#ifdef PROFILE
|
|
#define _Y pa_thread_yield()
|
|
#else
|
|
#define _Y do { } while(0)
|
|
#endif
|
|
|
|
struct pa_shmasyncq {
|
|
pa_fdsem *read_fdsem, *write_fdsem;
|
|
pa_shmasyncq_data *data;
|
|
};
|
|
|
|
static int is_power_of_two(unsigned size) {
|
|
return !(size & (size - 1));
|
|
}
|
|
|
|
static int reduce(pa_shmasyncq *l, int value) {
|
|
return value & (unsigned) (l->n_elements - 1);
|
|
}
|
|
|
|
static pa_atomic_t* get_cell(pa_shmasyncq *l, unsigned i) {
|
|
pa_assert(i < l->data->n_elements);
|
|
|
|
return (pa_atomic_t*) ((uint8*t) l->data + PA_ALIGN(sizeof(pa_shmasyncq_data)) + i * (PA_ALIGN(sizeof(pa_atomic_t)) + PA_ALIGN(element_size)))
|
|
}
|
|
|
|
static void *get_cell_data(pa_atomic_t *a) {
|
|
return (uint8_t*) a + PA_ALIGN(sizeof(atomic_t));
|
|
}
|
|
|
|
pa_shmasyncq *pa_shmasyncq_new(unsigned n_elements, size_t element_size, void *data, int fd[2]) {
|
|
pa_shmasyncq *l;
|
|
|
|
pa_assert(n_elements > 0);
|
|
pa_assert(is_power_of_two(n_elements));
|
|
pa_assert(element_size > 0);
|
|
pa_assert(data);
|
|
pa_assert(fd);
|
|
|
|
l = pa_xnew(pa_shmasyncq, 1);
|
|
|
|
l->data = data;
|
|
memset(data, 0, PA_SHMASYNCQ_SIZE(n_elements, element_size));
|
|
|
|
l->data->n_elements = n_elements;
|
|
l->data->element_size = element_size;
|
|
|
|
if (!(l->read_fdsem = pa_fdsem_new_shm(&d->read_fdsem_data))) {
|
|
pa_xfree(l);
|
|
return NULL;
|
|
}
|
|
fd[0] = pa_fdsem_get(l->read_fdsem);
|
|
|
|
if (!(l->write_fdsem = pa_fdsem_new(&d->write_fdsem_data, &fd[1]))) {
|
|
pa_fdsem_free(l->read_fdsem);
|
|
pa_xfree(l);
|
|
return NULL;
|
|
}
|
|
|
|
return l;
|
|
}
|
|
|
|
void pa_shmasyncq_free(pa_shmasyncq *l, pa_free_cb_t free_cb) {
|
|
pa_assert(l);
|
|
|
|
if (free_cb) {
|
|
void *p;
|
|
|
|
while ((p = pa_shmasyncq_pop(l, 0)))
|
|
free_cb(p);
|
|
}
|
|
|
|
pa_fdsem_free(l->read_fdsem);
|
|
pa_fdsem_free(l->write_fdsem);
|
|
pa_xfree(l);
|
|
}
|
|
|
|
int pa_shmasyncq_push(pa_shmasyncq*l, void *p, int wait) {
|
|
int idx;
|
|
pa_atomic_ptr_t *cells;
|
|
|
|
pa_assert(l);
|
|
pa_assert(p);
|
|
|
|
cells = PA_SHMASYNCQ_CELLS(l);
|
|
|
|
_Y;
|
|
idx = reduce(l, l->write_idx);
|
|
|
|
if (!pa_atomic_ptr_cmpxchg(&cells[idx], NULL, p)) {
|
|
|
|
if (!wait)
|
|
return -1;
|
|
|
|
/* pa_log("sleeping on push"); */
|
|
|
|
do {
|
|
pa_fdsem_wait(l->read_fdsem);
|
|
} while (!pa_atomic_ptr_cmpxchg(&cells[idx], NULL, p));
|
|
}
|
|
|
|
_Y;
|
|
l->write_idx++;
|
|
|
|
pa_fdsem_post(l->write_fdsem);
|
|
|
|
return 0;
|
|
}
|
|
|
|
void* pa_shmasyncq_pop(pa_shmasyncq*l, int wait) {
|
|
int idx;
|
|
void *ret;
|
|
pa_atomic_ptr_t *cells;
|
|
|
|
pa_assert(l);
|
|
|
|
cells = PA_SHMASYNCQ_CELLS(l);
|
|
|
|
_Y;
|
|
idx = reduce(l, l->read_idx);
|
|
|
|
if (!(ret = pa_atomic_ptr_load(&cells[idx]))) {
|
|
|
|
if (!wait)
|
|
return NULL;
|
|
|
|
/* pa_log("sleeping on pop"); */
|
|
|
|
do {
|
|
pa_fdsem_wait(l->write_fdsem);
|
|
} while (!(ret = pa_atomic_ptr_load(&cells[idx])));
|
|
}
|
|
|
|
pa_assert(ret);
|
|
|
|
/* Guaranteed to succeed if we only have a single reader */
|
|
pa_assert_se(pa_atomic_ptr_cmpxchg(&cells[idx], ret, NULL));
|
|
|
|
_Y;
|
|
l->read_idx++;
|
|
|
|
pa_fdsem_post(l->read_fdsem);
|
|
|
|
return ret;
|
|
}
|
|
|
|
int pa_shmasyncq_get_fd(pa_shmasyncq *q) {
|
|
pa_assert(q);
|
|
|
|
return pa_fdsem_get(q->write_fdsem);
|
|
}
|
|
|
|
int pa_shmasyncq_before_poll(pa_shmasyncq *l) {
|
|
int idx;
|
|
pa_atomic_ptr_t *cells;
|
|
|
|
pa_assert(l);
|
|
|
|
cells = PA_SHMASYNCQ_CELLS(l);
|
|
|
|
_Y;
|
|
idx = reduce(l, l->read_idx);
|
|
|
|
for (;;) {
|
|
if (pa_atomic_ptr_load(&cells[idx]))
|
|
return -1;
|
|
|
|
if (pa_fdsem_before_poll(l->write_fdsem) >= 0)
|
|
return 0;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
void pa_shmasyncq_after_poll(pa_shmasyncq *l) {
|
|
pa_assert(l);
|
|
|
|
pa_fdsem_after_poll(l->write_fdsem);
|
|
}
|