Lock-free implementation of a Thread Library

I have been going through a few published papers which details algorithms and data structures which can be used for implementation of a thread library. I searched online to view any existing implementation of the same, and I could only find LFThreads(only its documentation, not its source code).

Have any of you worked on/used a lock-free thread library? I would like to go through its source and a see a few examples of how exactly it is used.

From stackoverflow

• This doesn't exactly answer your question, but the following article might be helpful: http://www.ibm.com/developerworks/java/library/j-jtp11234/

• Have you looked at Intel Threading Building Blocks? It's C++, which is not in your tags, but I thought you might be interested since it's open source.