freebsd-dev/lib/libc/gen/sem_new.c
David Xu cc583082f4 Use version 2 of semaphore provided by kernel umtx code, now if there is
no waiters, we still increase and decrease count in user mode without
entering kernel, once there is a waiter, sem_post will enter kernel to
increase count and wake thread up, this is atomicy and allow us to
gracefully destroy semaphore after sem_wait returned.
2012-03-21 07:12:52 +00:00

469 lines
9.7 KiB
C

/*
* Copyright (C) 2010 David Xu <davidxu@freebsd.org>.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice(s), this list of conditions and the following disclaimer as
* the first lines of this file unmodified other than the possible
* addition of one or more copyright notices.
* 2. Redistributions in binary form must reproduce the above copyright
* notice(s), this list of conditions and the following disclaimer in
* the documentation and/or other materials provided with the
* distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER(S) ``AS IS'' AND ANY
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER(S) BE
* LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
* BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
* WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE
* OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
* EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*
* $FreeBSD$
*/
#include "namespace.h"
#include <sys/types.h>
#include <sys/queue.h>
#include <sys/mman.h>
#include <sys/stat.h>
#include <errno.h>
#include <machine/atomic.h>
#include <sys/umtx.h>
#include <limits.h>
#include <fcntl.h>
#include <pthread.h>
#include <stdarg.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <semaphore.h>
#include <unistd.h>
#include "un-namespace.h"
#include "libc_private.h"
__weak_reference(_sem_close, sem_close);
__weak_reference(_sem_destroy, sem_destroy);
__weak_reference(_sem_getvalue, sem_getvalue);
__weak_reference(_sem_init, sem_init);
__weak_reference(_sem_open, sem_open);
__weak_reference(_sem_post, sem_post);
__weak_reference(_sem_timedwait, sem_timedwait);
__weak_reference(_sem_trywait, sem_trywait);
__weak_reference(_sem_unlink, sem_unlink);
__weak_reference(_sem_wait, sem_wait);
#define SEM_PREFIX "/tmp/SEMD"
#define SEM_MAGIC1 ((u_int32_t)0x73656d31)
#define SEM_MAGIC ((u_int32_t)0x73656d32)
struct sem_nameinfo {
int open_count;
char *name;
sem_t *sem;
LIST_ENTRY(sem_nameinfo) next;
};
static pthread_once_t once = PTHREAD_ONCE_INIT;
static pthread_mutex_t sem_llock;
static LIST_HEAD(,sem_nameinfo) sem_list = LIST_HEAD_INITIALIZER(sem_list);
static void
sem_prefork()
{
_pthread_mutex_lock(&sem_llock);
}
static void
sem_postfork()
{
_pthread_mutex_unlock(&sem_llock);
}
static void
sem_child_postfork()
{
_pthread_mutex_unlock(&sem_llock);
}
static void
sem_module_init(void)
{
pthread_mutexattr_t ma;
_pthread_mutexattr_init(&ma);
_pthread_mutexattr_settype(&ma, PTHREAD_MUTEX_RECURSIVE);
_pthread_mutex_init(&sem_llock, &ma);
_pthread_mutexattr_destroy(&ma);
_pthread_atfork(sem_prefork, sem_postfork, sem_child_postfork);
}
static inline int
sem_check_validity(sem_t *sem)
{
if (sem->_magic == SEM_MAGIC || sem->_magic == SEM_MAGIC1)
return (0);
else {
errno = EINVAL;
return (-1);
}
}
int
_sem_init(sem_t *sem, int pshared, unsigned int value)
{
if (value > SEM_VALUE_MAX) {
errno = EINVAL;
return (-1);
}
bzero(sem, sizeof(sem_t));
sem->_magic = SEM_MAGIC;
sem->_kern._count = (u_int32_t)value;
sem->_kern._has_waiters = 0;
sem->_kern._flags = (pshared ? USYNC_PROCESS_SHARED : 0) | SEM_VER2;
return (0);
}
sem_t *
_sem_open(const char *name, int flags, ...)
{
char path[PATH_MAX];
struct stat sb;
va_list ap;
struct sem_nameinfo *ni = NULL;
sem_t *sem = NULL;
int fd = -1, mode, len, errsave;
int value = 0;
if (name[0] != '/') {
errno = EINVAL;
return (SEM_FAILED);
}
name++;
if (flags & ~(O_CREAT|O_EXCL)) {
errno = EINVAL;
return (SEM_FAILED);
}
_pthread_once(&once, sem_module_init);
_pthread_mutex_lock(&sem_llock);
LIST_FOREACH(ni, &sem_list, next) {
if (strcmp(name, ni->name) == 0) {
ni->open_count++;
sem = ni->sem;
_pthread_mutex_unlock(&sem_llock);
return (sem);
}
}
if (flags & O_CREAT) {
va_start(ap, flags);
mode = va_arg(ap, int);
value = va_arg(ap, int);
va_end(ap);
}
len = sizeof(*ni) + strlen(name) + 1;
ni = (struct sem_nameinfo *)malloc(len);
if (ni == NULL) {
errno = ENOSPC;
goto error;
}
ni->name = (char *)(ni+1);
strcpy(ni->name, name);
strcpy(path, SEM_PREFIX);
if (strlcat(path, name, sizeof(path)) >= sizeof(path)) {
errno = ENAMETOOLONG;
goto error;
}
fd = _open(path, flags|O_RDWR, mode);
if (fd == -1)
goto error;
if (flock(fd, LOCK_EX) == -1)
goto error;
if (_fstat(fd, &sb)) {
flock(fd, LOCK_UN);
goto error;
}
if (sb.st_size < sizeof(sem_t)) {
sem_t tmp;
tmp._magic = SEM_MAGIC;
tmp._kern._has_waiters = 0;
tmp._kern._count = value;
tmp._kern._flags = USYNC_PROCESS_SHARED | SEM_NAMED | SEM_VER2;
if (_write(fd, &tmp, sizeof(tmp)) != sizeof(tmp)) {
flock(fd, LOCK_UN);
goto error;
}
}
flock(fd, LOCK_UN);
sem = (sem_t *)mmap(NULL, sizeof(sem_t), PROT_READ|PROT_WRITE,
MAP_SHARED|MAP_NOSYNC, fd, 0);
if (sem == MAP_FAILED) {
sem = NULL;
if (errno == ENOMEM)
errno = ENOSPC;
goto error;
}
if (sem->_magic != SEM_MAGIC) {
errno = EINVAL;
goto error;
}
ni->open_count = 1;
ni->sem = sem;
LIST_INSERT_HEAD(&sem_list, ni, next);
_pthread_mutex_unlock(&sem_llock);
_close(fd);
return (sem);
error:
errsave = errno;
_pthread_mutex_unlock(&sem_llock);
if (fd != -1)
_close(fd);
if (sem != NULL)
munmap(sem, sizeof(sem_t));
free(ni);
errno = errsave;
return (SEM_FAILED);
}
int
_sem_close(sem_t *sem)
{
struct sem_nameinfo *ni;
if (sem_check_validity(sem) != 0)
return (-1);
if (!(sem->_kern._flags & SEM_NAMED)) {
errno = EINVAL;
return (-1);
}
_pthread_once(&once, sem_module_init);
_pthread_mutex_lock(&sem_llock);
LIST_FOREACH(ni, &sem_list, next) {
if (sem == ni->sem) {
if (--ni->open_count > 0) {
_pthread_mutex_unlock(&sem_llock);
return (0);
}
else
break;
}
}
if (ni) {
LIST_REMOVE(ni, next);
_pthread_mutex_unlock(&sem_llock);
munmap(sem, sizeof(*sem));
free(ni);
return (0);
}
_pthread_mutex_unlock(&sem_llock);
errno = EINVAL;
return (-1);
}
int
_sem_unlink(const char *name)
{
char path[PATH_MAX];
if (name[0] != '/') {
errno = ENOENT;
return -1;
}
name++;
strcpy(path, SEM_PREFIX);
if (strlcat(path, name, sizeof(path)) >= sizeof(path)) {
errno = ENAMETOOLONG;
return (-1);
}
return unlink(path);
}
int
_sem_destroy(sem_t *sem)
{
if (sem_check_validity(sem) != 0)
return (-1);
if (sem->_kern._flags & SEM_NAMED) {
errno = EINVAL;
return (-1);
}
sem->_magic = 0;
return (0);
}
int
_sem_getvalue(sem_t * __restrict sem, int * __restrict sval)
{
if (sem_check_validity(sem) != 0)
return (-1);
*sval = (int)sem->_kern._count & ~SEM_WAITERS;
return (0);
}
static __inline int
usem_wait(struct _usem *sem, const struct timespec *abstime)
{
struct _umtx_time *tm_p, timeout;
size_t tm_size;
if (abstime == NULL) {
tm_p = NULL;
tm_size = 0;
} else {
timeout._clockid = CLOCK_REALTIME;
timeout._flags = UMTX_ABSTIME;
timeout._timeout = *abstime;
tm_p = &timeout;
tm_size = sizeof(timeout);
}
return _umtx_op(sem, UMTX_OP_SEM_WAIT, 0,
(void *)tm_size, __DECONST(void*, tm_p));
}
static inline int
_trywait(sem_t *sem)
{
int count;
if ((sem->_kern._flags & SEM_VER2) != 0) {
while (((count = sem->_kern._count) & ~SEM_WAITERS) > 0) {
if (atomic_cmpset_acq_int(&sem->_kern._count, count, count - 1))
return (0);
}
} else {
while ((count = sem->_kern._count) > 0) {
if (atomic_cmpset_acq_int(&sem->_kern._count, count, count - 1))
return (0);
}
}
return (EAGAIN);
}
int
_sem_trywait(sem_t *sem)
{
int status;
if (sem_check_validity(sem) != 0)
return (-1);
if ((status = _trywait(sem)) == 0)
return (0);
errno = status;
return (-1);
}
int
_sem_timedwait(sem_t * __restrict sem,
const struct timespec * __restrict abstime)
{
int retval;
if (sem_check_validity(sem) != 0)
return (-1);
retval = 0;
for (;;) {
if (_trywait(sem) == 0)
return (0);
if (retval) {
_pthread_testcancel();
break;
}
/*
* The timeout argument is only supposed to
* be checked if the thread would have blocked.
*/
if (abstime != NULL) {
if (abstime->tv_nsec >= 1000000000 || abstime->tv_nsec < 0) {
errno = EINVAL;
return (-1);
}
}
_pthread_cancel_enter(1);
retval = usem_wait(&sem->_kern, abstime);
_pthread_cancel_leave(0);
}
return (retval);
}
int
_sem_wait(sem_t *sem)
{
return _sem_timedwait(sem, NULL);
}
/*
* POSIX:
* The sem_post() interface is reentrant with respect to signals and may be
* invoked from a signal-catching function.
* The implementation does not use lock, so it should be safe.
*/
int
_sem_post(sem_t *sem)
{
int count;
if (sem_check_validity(sem) != 0)
return (-1);
if ((sem->_kern._flags & SEM_VER2) != 0) {
for (;;) {
count = sem->_kern._count;
if ((count & SEM_WAITERS) == 0) {
if (__predict_false(count == SEM_VALUE_MAX)) {
errno = ERANGE;
return (-1);
}
if (atomic_cmpset_rel_int(&sem->_kern._count, count, count+1))
return (0);
} else {
return _umtx_op(&sem->_kern, UMTX_OP_SEM_WAKE, 0, NULL, NULL);
}
}
} else {
do {
count = sem->_kern._count;
if (__predict_false(count == SEM_VALUE_MAX)) {
errno = ERANGE;
return (-1);
}
} while (!atomic_cmpset_rel_int(&sem->_kern._count, count, count+1));
rmb();
if (!sem->_kern._has_waiters)
return (0);
return _umtx_op(&sem->_kern, UMTX_OP_SEM_WAKE, 0, NULL, NULL);
}
}