Implementing multithreading policy templates in D?

[Lars Ivar Igesund]

Brian Price wrote:

> Greetings,
> 
> While implementing a design I ran into a need for multiple implementations
> with different threading policies.  It looked to me as if porting Loki's
> Threading Model policies over to D would be just the ticket. 
> Unfortunately there's no mutex-like class in Phobos that I can find and
> Object's monitor is not exposed so no way to acquire/release on it other
> than through synchronized.
> 
> Without a mutex like object having aquire/release semantics, I'd have to
> litter my code with a bunch of compile time conditionals instead of using
> a RAII Lock object.  Such a Lock object could be provided via template
> argument allowing different threading policies for different
> instantiations.  In the 'non-thread-safe' scenario the Lock object would
> just be an empty object that (I assume) the compiler would optimize away.
> 
> Since I couldn't figure out how to make such a Lock object using the
> synchronized statement, I tried rolling my own mutex on top of D 2.014's
> standard library (Phobos) and language primitives.   Since volatile seems
> to be deprecated, my design uses two synchronization objects through
> synchronized statements and requires a waitable object that can be
> signaled by the releasing thread to avoid busy-wait.
> 
> At this point I hit brick wall #2 (#1 being the absence of a mutex).  The
> only waitable object I can find in the standard library is Thread and the
> only event you can wait on is its death.  Up to this point the pure D
> standard mutex implementation was doable (though undoubtedly less
> efficient than native implementations).  Creating a std.Thread derived
> class to be used as a one shot wait object pushed the design into the yes
> it will work but it's completely absurd camp.
> 
> Having used about every 'mainstream' language over the past twenty odd
> years, I figure either I'm missing something huge and need to learn an
> entirely new approach or there's something missing from the standard
> library.  So I'm left with three questions:
> 
> Did I miss something in the docs/std lib code?
> 
> Is there a way to implement flexible threading policies using synchronized
> statements?
> 
> What are the chances we'll see Object sporting wait/notify methods or
> lock/unlock methods in a future release?
> 
> Thanks,
> Brian

Tango should have most (if not all) the threading primitives you need - to
be frank, for any serious threading code, Phobos doesn't cut it.

-- 
Lars Ivar Igesund
blog at http://larsivi.net
DSource, #d.tango & #D: larsivi
Dancing the Tango