How I can solve the error "Data ... inferred as a variable size matrix" in Simulink?

I wrote a Matlab function in Simulink. The compilation process detects that u data is a variable size matrix, but it isn't on my case.

The code is showed below. The error is:

Data 'u' is inferred as a variable size matrix, while its properties in the Model Explorer specify its size as inherited or fixed. Please check the 'Variable Size' check box and specify the upper bounds in the size field.

But when I check the 'Variable Size' option, other problems emerge, because all of my signals presents fixed size.

function u = buffer(d_uf, arrival_time)
% ------------------------------------------------------------------------
% Parameters
m = 1;
Nc = 36;
% ------------------------------------------------------------------------
persistent index u_old d_uf_ready time_old

% Initialization of persistent variables
if isempty(u_old)
    u_old = zeros(m,1);
end
if isempty(d_uf_ready)
    d_uf_ready = zeros(Nc*m,1);
end
if isempty(index)
    index = 1;
end
if isempty(time_old)
    time_old = 0;
end

% Condition to use the buffer
if arrival_time < time_old
    d_u = d_uf_ready(index:index+m-1);
    index = index + m; % Move the index to the next block position
else % Update the buffer and use the most recent available data 
    d_u = d_uf(1:m);
    index = 1 + m;      % Restart the index counter
    d_uf_ready = d_uf;  % Storing to the next time sample
end
% Obtaining control signal
u = d_u + u_old;
% Storing to the next time sample
u_old = u;
end

[Victor Maciel]

I got a solution. The problem is on this code:

d_u = d_uf_ready(index:index+m-1);

index = index + m; % Move the index to the next block position

For some reason, Matlab understands that d_u maybe present a variable size. But it doesn't true, because the size is a fixed value m. My solution is:

d_u = d_uf_ready(1:m);

% Moving the block of available data to the next block position

dummy = d_uf_ready(m+1:end);

d_u = [dummy; zeros(m,1)];

Answer 1

[Walter Roberson]

u = d_u + u_old;

Before that assign

u = zeros(m, 1);

[Victor Maciel]

Thanks for answering. I got another solution, but I think that yours works.

[esmail mobasher]

hi victor, how did you solve?