Hi Everyone,

Crypto++ provides a Singleton class in misc.h. The original class suffered 
C++03 (and earlier) shortcomings. A design compromise was made to leak 
memory rather than using locks to increase portability.

We now have a portable implementation that does not depend on non-portable 
locks. It follows 
http://preshing.com/20130930/double-checked-locking-is-fixed-in-cpp11.

template <class T, class F, int instance>
  const T & Singleton<T, F, instance>::Ref(CRYPTOPP_NOINLINE_DOTDOTDOT) const
{
    static std::mutex s_mutex;
    static std::atomic<T*> s_pObject;

    T *p = s_pObject.load(std::memory_order_relaxed);
    std::atomic_thread_fence(std::memory_order_acquire);

    if (p)
        return *p;

    std::lock_guard<std::mutex> lock(s_mutex);
    p = s_pObject.load(std::memory_order_relaxed);
    std::atomic_thread_fence(std::memory_order_acquire);

    if (p)
        return *p;

    T *newObject = m_objectFactory();
    s_pObject.store(newObject, std::memory_order_relaxed);
    std::atomic_thread_fence(std::memory_order_release);

    return *newObject;
}

The implementation is guarded by both CRYPTOPP_CXX11_ATOMICS and 
CRYPTOPP_CXX11_SYNCHRONIZATION because atomics is not a proper subset of 
synchronization when it comes to compiler support and versions.

I'd like to merge it once it goes through a round of testing.

Are there any objections?

Jeff

-- 
-- 
You received this message because you are subscribed to the "Crypto++ Users" 
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at 
http://www.cryptopp.com.
--- 
You received this message because you are subscribed to the Google Groups 
"Crypto++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to