Re: Question on POSIX Condition Variables

[David Butenhof <david.butenhof@xxxxxx>]

Andrew Josey wrote:
> The following question was forwarded to me.Is there any further
> information  I could respond with other than just saying the former 1c
> threads extension?
>   
The direct ancestor is the DEC SRC Taos/Topaz work (a lot of people 
including Leslie Lamport, Andrew Birrell, and Garret Swart) on OS and 
language support for parallel processing in the mid-1980s. The SRC work 
was an important inspiration for the Digital "Common Multithreading 
Architecture" (CMA), which provided a base reference document for the 
first draft of Pthreads.

Essentially, the combination of mutex and condition variable is intended 
to provide efficient primitives to enable library-level support for 
something approximating Hoare monitors.

A good SRC reference to start further research would be "Synchronization 
Primitives for a Multiprocessor: A Formal Specification" (Birrell, 
Guttag, Horning, Levin).

  
http://www.hpl.hp.com/techreports/Compaq-DEC/SRC-RR-20.pdf?jumpid=reg_R1002_USEN
> regards
> Andrew
> ---------------
>
> I teach a course at Boston Univ on Operating Systems. I have become
> curious where the POSIX condition variables primitives came from.
> Who authored them? or what group did this particular approach come
> from?
>
> I realize that this is not exactly what you do, but I thought you
> could at least point me in the right direction.
>
> Thanks,
> John Day