This is an implementation of a M to N thread model. It mix the advantage of user-space thread and kernel-space thread. Differ from tranditional pth and pthread api, you can get dispatch your code to a seperate native pthread by simplely wrapping code with ht_hand_out() and ht_get_back() function calls.

example: //TBD