I concur with aragorn, solve this problem with select rather than threads. Question: How many processors does the computer you run your thread code on have? If it's just one, then the only way your code can achieve speed-up is if there is potential waiting due to I/O.

I too have experience with writing real threads with C code and the pthread.h library (see Solved N-Queens (warning: C code ahead). My computer at home is Dual processor and the computers i used at school had Dual and Quad processors. We saw true speed-up, even the occasional super-linear speedup when we threaded our programs. I think that a lot of people think that by threading your program, it just magically "speeds up." This is simply not true, especially if they are only using a One Proc Box (apologies to Sammy Hagar). Speed up happens on a One Proc Box when processes have to wait on things like on I/O, and select is a good (yet hard to understand) solution for this.

L-LL-L--L-LL-L--L-LL-L--
-R--R-RR-R--R-RR-R--R-RR
B--B--B--B--B--B--B--B--
H---H---H---H---H---H---
(the triplet paradiddle with high-hat)