SHARC C/C++ Code Optimization

I'm starting to learn about how to best optimize C/C++ code to run on the SHARC architecture (21369 and 21469 processors), and I was hoping some of might have some suggestions as to how to go about that.  For my project the plan is to mostly use C++ code and only use assembly where it's absolutely necessary.  I am also taking advantage of the DSP runtime library where applicable.

I read the "Achieving Optimal Performance from C/C++ Source Code" section in the VisualDSP Compiler User Guide, which was helpful, but I'm looking for more hands on examples or tutorials. 

In particular, I was wondering if any of you know of some C code examples which illustrates how to most efficiently implement a biquad (particularly direct form 2 transpose) on the SHARC?  Or any other code examples that might be helpful.  It would also be great to have code that includes benchmarks.

I found the 2136x application code library on the ADI website, but it appears to only have assembly code examples.

http://www.analog.com/en/processors-dsp/sharc/products/code-examples/2136x_application_code_examples/resources/fca.html

Any help will be appreciated.

Thanks,

Chris