Main Content

Generate Code by Using Embedded Coder Quick Start

ComponentDeploymentFcn is a simple export-function model that uses Initialize Function and Terminate Function blocks to represent startup and shutdown behavior and two function-call subsystems to represent the component algorithm. You prepare ComponentDeploymentFcn for embedded code generation by using Embedded Coder® Quick Start, which chooses code generation settings based on your goals and application.

Generate Code with Quick Start Tool

  1. Open the model ComponentDeploymentFcn by typing this command:

      'supportingFile', 'ComponentDeploymentFcn.slx');

    Export function model that includes an Initialize Function block, Terminate Function block, an triggered function-call subsystem, and a periodic function-call subsystem.

  2. If the C Code tab is not already open, in the Apps gallery, under Code Generation, click Embedded Coder.

  3. On the C Code tab, click Quick Start.

    Embedded Coder Quick Start display for model ComponentDeploymentFcn.

  4. Advance through the steps of the Quick Start tool, stopping at the Generate Code step. Each step asks questions about the code that you want to generate. For this tutorial, use the default settings that are already selected. The tool validates your selections against the model and presents the parameter changes required to generate code.

  5. In the Generate Code step, apply the proposed changes and generate code from ComponentDeploymentFcn by clicking Next.

    The Code Generation Complete page presents possible next steps and informs you that the tool configured the model for code generation that applies a service code interface. A service interface configuration maps model data elements to service interfaces that are defined and stored in a shared Embedded Coder Dictionary configured for the model.

  6. Click Finish.

  7. In the Embedded Coder app, return to the C Code tab. From this tab you can configure code generation customizations and check the results in the Code view pane.

    View of the ComponentDeploymentFcn model in context of Embedded Coder app. The toolstrip is at the top. The model is in the middle. The Code view pane is on the right.

Inspect the Generated Code

The generated code appears in three primary files: ComponentDeploymentFcn.c, ComponentDeploymentFcn.h, and services.h. In your MATLAB® current folder, the ComponentDeploymentFcn_ert_rtw folder contains the files ComponentDeploymentFcn.c and ComponentDeploymentFcn.h. The file services.h, which is the header file that specifies function prototypes for target platform services, is in folder services.

The folder slprj/ert/_sharedutils contains the file rtwtypes.h, which defines standard data types that the generated code uses by default. This folder contains generated files that can or must be shared between multiple models.

Files ComponentDeploymentFcn.h and ComponentDeploymentFcn.c declare and define entry-point functions, which you call from your application code. For this example model, the entry-point functions include an initialization function, two execution functions, and a terminate function. The functions exchange data with your target application code through service interfaces that you configure. The file services.h is a header file that declares interfaces that the generated entry-point functions use to call target platform services.

To see the entry-point functions in the generated code:

  1. On the right side of the Simulink® Editor window, in the Code view pane, locate the search bar.

  2. In the search bar, type CD_accumulator, the name of the entry-point function generated for the function-call subsystem Accumulator. To find instances of the function name across the generated code files, click the search suggestion.

    The string CD_integrator appears in the Code view search field and the response shows that Code view found an instance of the function name CD_integrator.

  3. In ComponentDeploymentFcn.c, inspect the code generated for entry-point function CD_accumulator.

    void CD_accumulator(void)
      const real_T *tmpIrvIRead;
      real_T rtb_Sum_0;
      int32_T i;
      tmpIrvIRead = get_CD_accumulator_DataTransfer();
      for (i = 0; i < 10; i++) {
        rtb_Sum_0 = tmpIrvIRead[i] + CD_measured.delay[i];
        (getref_CD_accumulator_OutBus_y())[i] = CD_tunable.k * rtb_Sum_0;
        CD_measured.delay[i] = rtb_Sum_0;

    The entry-point function calls the data transfer service function get_CD_accumulator_DataTransfer to read the output value transferred from the integrator function. For each element of the bus signal, the function applies a delay and gain value and writes the output by calling target platform sender service function getref_CD_accumulator_OutBus_y. The data communication method and function names that the code generator applies for the service calls are defined by the service interfaces selected in the Code Mappings editor. By default, the code mappings specify service interfaces configured for data communication that occurs outside of (before and after) function execution and applies data concurrency safeguards. The values of the input data remain unchanged as the function executes. Each time the function accesses the data, the code uses the same value. This method of data communication favors memory optimization over data freshness.

  4. In the Code view, use the arrows on the right to step through each instance of CD_accumulator, including the function definition in ComponentDeploymentFcn.c and the declaration in ComponentDeploymentFcn.h. You can also see the number of search hits in each file from the file menu in the upper left corner.

    Code view display of search hit count for each generated code file.

  5. Repeat the search steps to locate and explore instances of the other entry-point functions.

Next, change the service interface configuration and review differences in the generated code.