SYNOPSIS

#include <pthread.h>

int pthread_mutexattr_destroy(pthread_mutexattr_t *attr);
int pthread_mutexattr_init(pthread_mutexattr_t *attr);

DESCRIPTION

The pthread_mutexattr_destroy() function shall destroy a mutex attributes object; the object becomes, in effect, uninitialized. An implementation may cause pthread_mutexattr_destroy() to set the object referenced by attr to an invalid value. A destroyed attr attributes object can be reinitialized using pthread_mutexattr_init(); the results of otherwise referencing the object after it has been destroyed are undefined.

The pthread_mutexattr_init() function shall initialize a mutex attributes object attr with the default value for all of the attributes defined by the implementation.

Results are undefined if pthread_mutexattr_init() is called specifying an already initialized attr attributes object.

After a mutex attributes object has been used to initialize one or more mutexes, any function affecting the attributes object (including destruction) shall not affect any previously initialized mutexes.

RETURN VALUE

Upon successful completion, pthread_mutexattr_destroy() and pthread_mutexattr_init() shall return zero; otherwise, an error number shall be returned to indicate the error.

EXAMPLES

None.

APPLICATION USAGE

None.

RATIONALE

See pthread_attr_init() for a general explanation of attributes. Attributes objects allow implementations to experiment with useful extensions and permit extension of this volume of IEEE Std 1003.1-2001 without changing the existing functions. Thus, they provide for future extensibility of this volume of IEEE Std 1003.1-2001 and reduce the temptation to standardize prematurely on semantics that are not yet widely implemented or understood.

Examples of possible additional mutex attributes that have been discussed are spin_only, limited_spin, no_spin, recursive, and metered. (To explain what the latter attributes might mean: recursive mutexes would allow for multiple re-locking by the current owner; metered mutexes would transparently keep records of queue length, wait time, and so on.) Since there is not yet wide agreement on the usefulness of these resulting from shared implementation and usage experience, they are not yet specified in this volume of IEEE Std 1003.1-2001. Mutex attributes objects, however, make it possible to test out these concepts for possible standardization at a later time.

Mutex Attributes and Performance

Care has been taken to ensure that the default values of the mutex attributes have been defined such that mutexes initialized with the defaults have simple enough semantics so that the locking and unlocking can be done with the equivalent of a test-and-set instruction (plus possibly a few other basic instructions).

There is at least one implementation method that can be used to reduce the cost of testing at lock-time if a mutex has non-default attributes. One such method that an implementation can employ (and this can be made fully transparent to fully conforming POSIX applications) is to secretly pre-lock any mutexes that are initialized to non-default attributes. Any later attempt to lock such a mutex causes the implementation to branch to the "slow path" as if the mutex were unavailable; then, on the slow path, the implementation can do the "real work" to lock a non-default mutex. The underlying unlock operation is more complicated since the implementation never really wants to release the pre-lock on this kind of mutex. This illustrates that, depending on the hardware, there may be certain optimizations that can be used so that whatever mutex attributes are considered "most frequently used" can be processed most efficiently.

Process Shared Memory and Synchronization

The existence of memory mapping functions in this volume of IEEE Std 1003.1-2001 leads to the possibility that an application may allocate the synchronization objects from this section in memory that is accessed by multiple processes (and therefore, by threads of multiple processes).

In order to permit such usage, while at the same time keeping the usual case (that is, usage within a single process) efficient, a process-shared option has been defined.

If an implementation supports the _POSIX_THREAD_PROCESS_SHARED option, then the process-shared attribute can be used to indicate that mutexes or condition variables may be accessed by threads of multiple processes.

The default setting of PTHREAD_PROCESS_PRIVATE has been chosen for the process-shared attribute so that the most efficient forms of these synchronization objects are created by default.

Synchronization variables that are initialized with the PTHREAD_PROCESS_PRIVATE process-shared attribute may only be operated on by threads in the process that initialized them. Synchronization variables that are initialized with the PTHREAD_PROCESS_SHARED process-shared attribute may be operated on by any thread in any process that has access to it. In particular, these processes may exist beyond the lifetime of the initializing process. For example, the following code implements a simple counting semaphore in a mapped file that may be used by many processes.

/* sem.h */
struct semaphore {
    pthread_mutex_t lock;
    pthread_cond_t nonzero;
    unsigned count;
};
typedef struct semaphore semaphore_t;

semaphore_t *semaphore_create(char *semaphore_name);
semaphore_t *semaphore_open(char *semaphore_name);
void semaphore_post(semaphore_t *semap);
void semaphore_wait(semaphore_t *semap);
void semaphore_close(semaphore_t *semap);

/* sem.c */
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/mman.h>
#include <fcntl.h>
#include <pthread.h>
#include "sem.h"

semaphore_t *
semaphore_create(char *semaphore_name)
{
int fd;
    semaphore_t *semap;
    pthread_mutexattr_t psharedm;
    pthread_condattr_t psharedc;

    fd = open(semaphore_name, O_RDWR | O_CREAT | O_EXCL, 0666);
    if (fd < 0)
        return (NULL);
    (void) ftruncate(fd, sizeof(semaphore_t));
    (void) pthread_mutexattr_init(&psharedm);
    (void) pthread_mutexattr_setpshared(&psharedm,
        PTHREAD_PROCESS_SHARED);
    (void) pthread_condattr_init(&psharedc);
    (void) pthread_condattr_setpshared(&psharedc,
        PTHREAD_PROCESS_SHARED);
    semap = (semaphore_t *) mmap(NULL, sizeof(semaphore_t),
            PROT_READ | PROT_WRITE, MAP_SHARED,
            fd, 0);
    close (fd);
    (void) pthread_mutex_init(&semap->lock, &psharedm);
    (void) pthread_cond_init(&semap->nonzero, &psharedc);
    semap->count = 0;
    return (semap);
}

semaphore_t *
semaphore_open(char *semaphore_name)
{
    int fd;
    semaphore_t *semap;

    fd = open(semaphore_name, O_RDWR, 0666);
    if (fd < 0)
        return (NULL);
    semap = (semaphore_t *) mmap(NULL, sizeof(semaphore_t),
            PROT_READ | PROT_WRITE, MAP_SHARED,
            fd, 0);
    close (fd);
    return (semap);
}

void
semaphore_post(semaphore_t *semap)
{
    pthread_mutex_lock(&semap->lock);
    if (semap->count == 0)
        pthread_cond_signal(&semapx->nonzero);
    semap->count++;
    pthread_mutex_unlock(&semap->lock);
}

void
semaphore_wait(semaphore_t *semap)
{
    pthread_mutex_lock(&semap->lock);
    while (semap->count == 0)
        pthread_cond_wait(&semap->nonzero, &semap->lock);
    semap->count--;
    pthread_mutex_unlock(&semap->lock);
}

void
semaphore_close(semaphore_t *semap)
{
    munmap((void *) semap, sizeof(semaphore_t));
}

The following code is for three separate processes that create, post, and wait on a semaphore in the file /tmp/semaphore. Once the file is created, the post and wait programs increment and decrement the counting semaphore (waiting and waking as required) even though they did not initialize the semaphore.

/* create.c */
#include "pthread.h"
#include "sem.h"

int
main()
{
    semaphore_t *semap;

    semap = semaphore_create("/tmp/semaphore");
    if (semap == NULL)
        exit(1);
    semaphore_close(semap);
    return (0);
}

/* post */
#include "pthread.h"
#include "sem.h"

int
main()
{
    semaphore_t *semap;

    semap = semaphore_open("/tmp/semaphore");
    if (semap == NULL)
        exit(1);
    semaphore_post(semap);
    semaphore_close(semap);
    return (0);
}

/* wait */
#include "pthread.h"
#include "sem.h"

int
main()
{
    semaphore_t *semap;

    semap = semaphore_open("/tmp/semaphore");
    if (semap == NULL)
        exit(1);
    semaphore_wait(semap);
    semaphore_close(semap);
    return (0);
}

FUTURE DIRECTIONS

None.

COPYRIGHT

Portions of this text are reprinted and reproduced in electronic form from IEEE Std 1003.1, 2003 Edition, Standard for Information Technology -- Portable Operating System Interface (POSIX), The Open Group Base Specifications Issue 6, Copyright (C) 2001-2003 by the Institute of Electrical and Electronics Engineers, Inc and The Open Group. In the event of any discrepancy between this version and the original IEEE and The Open Group Standard, the original IEEE and The Open Group Standard is the referee document. The original Standard can be obtained online at http://www.opengroup.org/unix/online.html .