I want use the CMSIS-DSP library to calculate the real FFT of a signal, but when I call the FFT functions and I compile in CodeWarrior, the code size exceeds the flash memory.
I´m using a MK20DX128VLH5 with 128kB of flash and my code without FFT function has 6kB.
FFT functions of the library are so great right? Is it correct this?
Don´t have a reduce form of FFT functions to use with Kinetis?
Hi Vinicius
Any update regarding implementation of the library?
Regards
Amit Kumar
Hi Amit Kumar
No, I gave up on implement the library. The code with CMSIS-DSP FFT is very big and I am not able to reduce it.
Regards
Vinícius Hoff
Hi Vinicius
I figured a way how to implement CMSIS lib in Codewarrior please refer to the following post Interfacing CMSIS Library in Codewarrior? You don't need to impement the entire library as it is really huge. but you can implement the CMSIS functions and related lib in your project. Hope It helps
Best Regards
Amit Kumar
Thank you Amit Kumar,
I'll try to implement the library again.
I´m using the library file "libarm_cortexM4l_math.a", available in the folder /Lib/GCC. There are a "arm_cortexM4l_math.lib" in the folder /Lib/ARM.
What´s the difference between them?
The "libarm_cortexM4lf_math.a" library is for mcu's with a hardware FPU. The "libarm_cortexM4l_math.a" library is for ones without hardware FPU.