If you're writing modern OpenGL, you know that tons of the built in matrix functions have been deprecated, and aren't available. It's not the worst thing in the world, because most of the math functions are well documented in multiple places, it's just a slight annoyance.
I happened to notice ZIP, from Lighthouse3D, which has the file vsml.cpp that gives the math for most of the common gl matrix functions (ortho, frustum, lookAt, etc.).
I also happen to have recently placed some glsl shaders in the drop box that do math of some of these matrix type of functions in the vertex shader portion, from thinking about this very issue. The file in the aforementioned zip, vsml.cpp, is probably a handy reference though. (Caveat, I haven't personally used the math in the file... double check your results!).
This is a more complete reference/alternate attempt - http://glm.g-truc.net/ that is also well worthwhile.