Início / user-19318975

hyunseo welcome's questions

Martin Hope
hyunseo welcome
Asked: 2025-04-08 03:02:55 +0800 CST
  • 6

Na documentação do Open Group , afirma-se que:

NAME
pthread_rwlock_wrlock, pthread_rwlock_trywrlock - lock a read-write lock object for writing

...

The pthread_rwlock_wrlock() and pthread_rwlock_trywrlock() functions may fail if:

...

[EDEADLK]
The current thread already owns the read-write lock for writing or reading.

Entretanto, quando executo o seguinte código:

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <errno.h>
#include <pthread.h>

int main()
{
    int error; 

    pthread_rwlock_t *mutex = (pthread_rwlock_t *)malloc(sizeof(pthread_rwlock_t));
    if (mutex == NULL) {
        perror("malloc");
        abort();
    }

    pthread_rwlockattr_t attr;

    error = pthread_rwlockattr_init(&attr);
    if (error != 0)
    {
        fprintf(stderr, "%s:%d: pthread_rwlockattr_init failed: %s", __FILE__, __LINE__, strerror(error));
        abort();
    }

    error = pthread_rwlock_init(mutex, &attr);
    if (error != 0)
    {
        fprintf(stderr, "%s:%d: pthread_rwlock_init failed: %s", __FILE__, __LINE__, strerror(error));
        abort();
    }
    
    error = pthread_rwlock_wrlock(mutex);
    if (error != 0)
    {
        fprintf(stderr, "%s:%d: pthread_rwlock_wrlock failed: %s", __FILE__, __LINE__, strerror(error));
        abort();
    }
    
    error = pthread_rwlock_rdlock(mutex);
    if (error != 0)
    {
        fprintf(stderr, "%s:%d: pthread_rwlock_rdlock failed: %s", __FILE__, __LINE__, strerror(error));
        abort();
    }

    return 0;
}

A saída é:

main.c:53: pthread_rwlock_rdlock failed: Resource deadlock avoided

Mas quando executo o seguinte código:

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <errno.h>
#include <pthread.h>

int main()
{
    int error; 

    pthread_rwlock_t *mutex = (pthread_rwlock_t *)malloc(sizeof(pthread_rwlock_t));
    if (mutex == NULL) {
        perror("malloc");
        abort();
    }

    pthread_rwlockattr_t attr;

    error = pthread_rwlockattr_init(&attr);
    if (error != 0)
    {
        fprintf(stderr, "%s:%d: pthread_rwlockattr_init failed: %s", __FILE__, __LINE__, strerror(error));
        abort();
    }

    error = pthread_rwlock_init(mutex, &attr);
    if (error != 0)
    {
        fprintf(stderr, "%s:%d: pthread_rwlock_init failed: %s", __FILE__, __LINE__, strerror(error));
        abort();
    }
    
    error = pthread_rwlock_rdlock(mutex);
    if (error != 0)
    {
        fprintf(stderr, "%s:%d: pthread_rwlock_rdlock failed: %s", __FILE__, __LINE__, strerror(error));
        abort();
    }
    
    error = pthread_rwlock_wrlock(mutex);
    if (error != 0)
    {
        fprintf(stderr, "%s:%d: pthread_rwlock_wrlock failed: %s", __FILE__, __LINE__, strerror(error));
        abort();
    }
        
    return 0;
}

Nesse caso, a pthread_rwlock_wrlockchamada não retorna EDEADLKe é bloqueada.

A explicação do Open Group está incorreta? Parece mais razoável que EDEADLKseja devolvida. Qual poderia ser o problema aqui?

Testei isso no Linux usando o gcc, no meu computador com o WSL e também no OnlineGDB, e não funciona em nenhum desses ambientes. Parece ser um problema comum de implementação, e não um problema específico da minha configuração. Esse comportamento é intencional?

c