Fw: [Boost-users] Problem with shared_ptr and pthreads_mutex_lock

[Paolo Carlini]

Paolo Carlini wrote:

>Indeed, the mutex is not initialized at all. The below is a possible,
>minimal, fix (for sure we are a long way from a well tested and
>optimized implementation). Can you test it (*) with your application??
>  
>
Unfortunately the draft patch would not work, in general, because we
would end up with different mutexes in different translation units... I
came to this conclusion this morning and, at the same time, Jonathan
sent me a message coming to the same conclusion.

Paolo.