Re: [boost] win32 mutex and condition variable algorithms (was:[threads] Permission given to change to Boost.License)

From: Peter Dimov (pdimov_at_[hidden])
Date: 2006-09-19 11:42:05

Anthony Williams wrote:

> The mutex is simple, and you can see the code in
> boost/thread/win32/basic_mutex.hpp on the thread_rewrite branch.
>
> The algorithm is:
>
> init():
>   active_count=0;
>   no semaphore
>
> lock():
>   atomic increment active_count
>   if new active_count ==1, that's us, so we've got the lock
>   else
>        get semaphore, and wait
>        now we've got the lock
>
> unlock():
>   atomic decrement active_count
>   if new active_count >0, then other threads are waiting,
>       so release semaphore.

This is what CRITICAL_SECTIONs do, and I recall being told by Alexander 
Terekhov that it is suboptimal (threads take the slow path when they could 
take the fast path). My tests with a contended queue seemed to prove him 
right.

Unfortunately, I can no longer find his post where he explained that with 
Google.