Re: Weak semaphores in Boost threads library

From: williamkempf_at_[hidden]
Date: 2001-10-17 12:23:11

--- In boost_at_y..., kevin_vanhorn_at_n... wrote:
> > I just noticed, to my dismay, that the thread library only
> > implements weak semaphores.

Not precisely true.  The Win32 implementation isn't a "weak" 
implementation.  The POSIX one may well be, however.  I'll 
investigate.

> I believe the following is a correct implementation of strong, fair
> semaphores built on top of POSIX mutexes and condition variables.
> It does depend on the assumption that POSIX mutexes are fair -- that
> is, a thread blocked on a mutex cannot remain forever blocked as 
long
> as every thread that locks the mutex eventually unlocks it.
> 
> class semaphore
> {
>   pthread_mutex_t mutex;
>   pthread_cond_t signal_occurred;
>   unsigned sem_cnt;
>   std::queue<bool *, std::list<bool *> > blocked_list;
> public:
>   semaphore(unsigned cnt);
>   
>   void signal()
>   {
>     pthread_mutex_lock(&mutex);
>     if (blocked_list.empty()) {
>       ++sem_cnt;
>     }
>     else {
>       *(blocked_list.front()) = false;
>       blocked_list.pop();
>       pthread_cond_broadcast(&signal_occurred);
>     }
>     pthread_mutex_unlock(&mutex);
>   }
> 
> 
>   void wait()
>   {
>     pthread_mutex_lock(&mutex);
>     if (sem_cnt > 0) {
>       --sem_cnt;
>     }
>     else {
>       bool blocked = true;
>       blocked_list.push(&blocked);
>       while (blocked)
>         pthread_cond_wait(&signal_occurred, &mutex);
>     }
>     pthread_mutex_unlock(&mutex);
>   }
> 
>   ~semaphore()
>   {
>     pthread_mutex_destroy(&mutex);
>     pthread_cond_destroy(&signal_occurred);
>   }
> };

Thanks for the implementation.  It will help in the evaluation.

Bill Kempf