Why do i get an error in the Fixed Point Conversion Step?

Error: The function 'HDLIFFT128' contains persistent variables 'ifft128' and has specialization 'HDLIFFT128_s1' associated with it. Function specializations containing persistent variables are not supported for fixed-point conversion. Consider rewriting your algorithm such that it does not use persistent variables within functions that require specialization.

I am trying to generate HDL Code from this main function.

 function [out1,out2] = main(x,y,z)
 %call subfunction 1
out1 = sub1(x);
 %call subfunction 2
out2 = sub2(y,z);
end

Sub-function 2 code:

 function [yOut,validOut] = HDLIFFT128(yIn,validIn)
%HDLIFFT128 
% Processes one sample of data using the dsp.HDLIFFT System object(TM)
% yIn is a fixed-point scalar or column vector. 
% validIn is a logical scalar.
% You can generate HDL code from this function.

persistent ifft128;
if isempty(ifft128)
  ifft128 = dsp.HDLIFFT('FFTLength',128);
end    
[yOut,validOut] = step(ifft128,yIn,validIn);
end

Answer 1

[Walter Roberson]

I would suggest breaking it up into two pieces, one of which constructs the ifft128 object, and the other of which expects that object to be always input.

[Pablo Medina]

Do You mean breaking the HDLIFFT128 fuction up? Could you please explain better your answer. Thanks for helping!

[Walter Roberson]

function ifft128 = init_HDLIFFT128
  ifft128 = dsp.HDLIFFT('FFTLength',128);

function [yOut,validOut] = HDLIFFT128(yIn, validIn, ifft128)
  [yOut,validOut] = step(ifft128,yIn,validIn);

The init function would be outside of any loop, and you would route its output to be an input of HDLIFFT128 that might be in a loop.

[Pablo Medina]

Thanks that solved the fixed point generation error.

Answer 2

[Bharath Venkataraman]

Pablo, splitting the function into a main function and calling the IFFT function from it resulted in HDL code generation. Here are the functions split up. Please replace the testbench and design files as needed.