C674x efficient FFT

Hello, I'm implementing FFT on the C6742 dsp, and performance is very critical.

I read carefully:

[1] http://processors.wiki.ti.com/index.php/Efficient_FFT_Computation_of_Real_Input

[2] http://www.ti.com/lit/an/spra291/spra291.pdf

And experimenting now with "Computing a Length N/2 Complex FFT from a Length N Real Input Sequence", using C example project from [1].

However I'd like to use ASM implementation, [2] contains it for C62xx, so where I can find for C674x? <-- this is the question.