When I´m writing shaders one of my biggest problems is that (because of the lack of intellisense) I don´t know what I can, and what I can´t do. This is specially a problem for basic maths. I recently found out that normalize exists in shaders (I was doing it manually before) but what about other methods? Is there a reference with all the available methods? what about lerp, dot product, cross product…
I think I found it:
http://http.developer.nvidia.com/Cg/index_stdlib.html
The Cg Tutorial - Appendix E. Cg Standard Library Functions (Same as above but with nice pictures).