Generate C/C++ Code - MATLAB & Simulink (original) (raw)

Main Content

After you prepare your MATLAB® code for code generation, specify input types, and check your code for issues, you can generate standalone C/C++ code or an accelerated MEX function. See theCode Generation Workflow.

Generate Standalone C/C++ Code

To generate standalone C/C++ code:

• In the MATLAB Coder™ app, from the Output Type list on theMATLAB Coder toolstrip, select Static Library (.lib),Dynamic Library (.dll), or Executable (.exe). Then, click the Generate Code button.
• At the command line, pass your entry-point functions to thecodegen function and use the -config option to specify a standalone C/C++ output type. See Configure Code Generation and Build Settings.

After you generate standalone C/C++ code, you can:

• Create a standalone C/C++ executable and deploy this application on a target platform.
• Integrate the generated source code or library into an existing C/C++ project.
• Run the generated code on embedded microprocessors by using a toolchain.
• Package the generated code for transfer to another development environment.

For details, see Deploy Generated Code and Package Code for Other Development Environments.

Generate an Accelerated MEX Function

To generate an accelerated MEX function:

• In the MATLAB Coder app, from the Output Type list on theMATLAB Coder toolstrip, select MEX. Then, click theGenerate Code button.
• At the command line, pass your entry-point functions to thecodegen function and use the -config option to specify MEX output. See Configure Code Generation and Build Settings.

Before you generate an accelerated MEX function:

• Review the available optimization strategies. In particular, consider unrolling for-loops and parfor-loops, inlining small functions, and passing variables by reference. See Optimize Generated C/C++ and MEX Code.
• Disable run-time checks. You should only disable run-time checks if you have checked your MATLAB code for issues by following the Code Generation Workflow. To disable run-time checks, see Control Run-Time Checks.

After you generate an accelerated MEX, you can call the MEX function in MATLAB. See Call MEX Functions. You can also explore further opportunities for acceleration by using the MATLAB Profiler to identify portions of the generated MEX function that are highly time-intensive. See Profile MEX Functions by Using MATLAB Profiler.

Accelerated MEX functions can be used to speed applications and simulations in a number of different toolboxes. For example:

• Real-time audio. See Accelerate Fixed-Point Simulation (Fixed-Point Designer).
• Pixel-streaming video processing. See Accelerate Pixel-Streaming Designs Using MATLAB Coder (HDL Coder).
• Communications algorithm simulation. See Simulation Acceleration Using MATLAB Coder and Parallel Computing Toolbox (Communications Toolbox).
• Standalone application deployment. See Use Generated Code to Accelerate an Application Deployed with MATLAB Compiler.

See Also

codegen

Topics

• Use an Example C Main in an Application
• Using Dynamic Memory Allocation for an Atoms Simulation
• Generate MEX Code to Accelerate Simulation of Bouncing Balls
• SIL and PIL Verification for Deployment on Raspberry Pi (Embedded Coder)
• Overview of Code Generation Using MATLAB Coder
• C++ Code Generation
• Tracing Generated C/C++ Code to MATLAB Source Code
• Best Practices for Using MEX Functions to Accelerate MATLAB Algorithms

External Websites

• Accelerating MATLAB Algorithms and Applications