Simulink Mathlab Function Block Output Variable Size Array

Hi,

I encountered the problem about array size from matlab function block. First of all I was creating a simulink project about data control. Its include some input values (constant blocks), mathlab function and display block. The input data is the address value, the length of the data to be transmitted and other values. Also, all the value type is uint8. Matlab function block is doing generate new array for output data. This array length is variable (regValue). For example lets assume the regValue is a 3. newArray length is a equal to that. In that case newArray = [data1 data2 data3]. But this program getting failure about variable type. At the same time, pins y5 and y6 are connected to the display block. How can I do this?

Some Errors;

Block Diagram:

Code:

function [y1, y2, y3, y4, y5] = response(slaveAddr, funcCode, startAddr, regValue)
    %#codegen
    % slaveAddr: 1 byte (Slave adresi)
    % funcCode: 1 byte (Fonksiyon kodu)
    % startAddr: 2 byte (Başlangıç adresi)
    % regValue: 2 byte (Register değeri veya değerleri)
    % crc1, crc2: 1 byte (Her iki CRC byte'ı)
    % generalData: (23 byte veri)

    % Verileri başlatıyoruz
    y1 = uint8(slaveAddr);          % 1 byte slave address
    y2 = uint8(funcCode);           % 1 byte function code
    y3 = uint8(regValue * 2);       % 1 byte Byte Count
    totalValue = (y3+2);                  % Başlangıçta boş veri High
    y4 = zeros(1, totalValue);
    y5 = zeros(1, length(y3));
 
end

Answer 1

[Walter Roberson]

Insert

totalValue = uint8(0);

before

totalValue = (y3+2);

[Semih]

Its correct but if you want to getting y4 or y5 value to outside getting failure about size. I want to solve this.

[Walter Roberson]

y3 = uint8(regValue * 2); 

That is a single uint8

y5 = zeros(1, length(y3));

length() of a single uint8 is always going to be 1. It would be clearer if you just coded

y5 = 0;

[Walter Roberson]

    y3 = uint8(regValue * 2);       % 1 byte Byte Count
    totalValue = (y3+2);                  % Başlangıçta boş veri High
    y4 = zeros(1, totalValue);

y3 is a uint8, so the maximum value it can hold is 255.

Adding 2 to a uint8 with value 255, gets 255 again. So the maximum value of totalValue is 255.

It follows that y4 is somewhere between 2 elements and 255 elements.

You need to code y4 first using coder.varsize giving an appropriate upperbound (at most 255)

[Semih]

Hi Walter,

It was solved. I done that you said. By adding the "%#codegen" and "coder.varsize('Y', [1, 100]);" code solve. Also I change the output parameters settings from Edit Data. This changing, size is a [1,100] and "Advanced Variable size" is marked. And I used this blok in enabled subsystem. The enabled system block was getting failure. Change the "Propagate size of variable-size signals:". I used "During execution" instead of "Only when enabling". Thank you for your help.