Use precompiled libraries

Whether to use precompiled libraries in generated code

Since R2024b

Description

App Configuration Pane: Custom Code

Configuration Objects: coder.MexCodeConfig | coder.CodeConfig | coder.EmbeddedCodeConfig

The Use precompiled libraries parameter specifies the extent to which the generated code uses platform-specific precompiled libraries.

If you want to optimize the performance of the generated code for a specific platform, instruct the code generator to prefer to use precompiled libraries when available.

If you want to create portable applications that can run on many platforms, instruct the code generator to avoid using precompiled libraries if alternative implementations of their algorithms are available.

For certain precompiled libraries (for example, BLAS, LAPACK, and FFTW), there exist individual configuration parameters that allow you to customize their usage. The Use precompiled libraries parameter does not affect the use of these libraries.

Settings

Prefer precompiled libraries when available

The code generator prefers to use the available platform-specific precompiled libraries. For C/C++ code generation, this is the default value.

Avoid precompiled libraries when possible

The code generator uses platform-specific precompiled libraries only if no alternative implementations of their algorithms are available. For CUDA® code generation (requires GPU Coder™), this is the default value.

Programmatic Use

Property: UsePrecompiledLibraries
Values: 'Prefer' | 'Avoid'
Default: 'Prefer'

Examples

Generate Portable Code for Image Filtering Algorithm

This example shows how you can use the UsePrecompiledLibraries code configuration property to generate portable code for an image processing function. This example requires the Image Processing Toolbox™.

  1. Create a MATLAB® entry-point function libraryUsageTest that uses the imfilter (Image Processing Toolbox) function to filter an input image.

    function filteredImage = libraryUsageTest(unfilteredImage)
h = [-1 0 1];
filteredImage = imfilter(unfilteredImage,h);
end

  2. Generate a static C library for the libraryUsageTest function. Specify a code configuration object with the InstructionSetExtensions property set to 'None'. This setting disables the generation of hardware-specific single instruction, multiple data (SIMD) code. Specify the type of the input argument to be an 8-by-8 array of doubles.

    cfg = coder.config("lib");
cfg.InstructionSetExtensions = "None";
codegen -config cfg libraryUsageTest -args {zeros(8)} -report

    Inspect the generated C function imfilter. This last line of this function contains a call to the precompiled library ippfilter_real64.

    void imfilter(const double varargin_1[64], double b[64])
{
  double aTmp[80];
  int i;
  for (i = 0; i < 8; i++) {
    aTmp[i] = 0.0;
    aTmp[i + 72] = 0.0;
    memcpy(&aTmp[i * 8 + 8], &varargin_1[i * 8], 8U * sizeof(double));
  }
  double kernel[3];
  double kernelSizeT[2];
  double outSizeT[2];
  double padSizeT[2];
  outSizeT[0] = 8.0;
  padSizeT[0] = 8.0;
  outSizeT[1] = 8.0;
  padSizeT[1] = 10.0;
  kernel[0] = -1.0;
  kernel[1] = 0.0;
  kernel[2] = 1.0;
  kernelSizeT[0] = 1.0;
  kernelSizeT[1] = 3.0;
  ippfilter_real64(&aTmp[0], &b[0], &outSizeT[0], 2.0, &padSizeT[0], &kernel[0],
                   &kernelSizeT[0], false);
}

  3. To generate code that contains a portable implementation of the image filtering algorithm as C source code, in the code configuration object cfg, set the UsePrecompiledLibraries property to "Avoid".

    cfg.UsePrecompiledLibraries = "Avoid";
codegen -config cfg libraryUsageTest -args {zeros(8)} -report

    Inspect the generated C function imfilter. This function does not contain calls to precompiled libraries.

    void imfilter(const double varargin_1[64], double b[64])
{
  double a_padded[80];
  int b_i;
  int i;
  int jb;
  for (i = 0; i < 8; i++) {
    a_padded[i] = 0.0;
    a_padded[i + 72] = 0.0;
    memcpy(&a_padded[i * 8 + 8], &varargin_1[i * 8], 8U * sizeof(double));
  }
  memset(&b[0], 0, 64U * sizeof(double));
  for (i = 0; i < 8; i++) {
    int cColOffset;
    cColOffset = i << 3;
    for (jb = 0; jb < 3; jb++) {
      int ia;
      ia = (i + jb) << 3;
      for (b_i = 0; b_i < 8; b_i++) {
        int b_tmp;
        b_tmp = cColOffset + b_i;
        b[b_tmp] += (-(2.0 - (double)jb) + 1.0) * a_padded[ia + b_i];
      }
    }
  }
}

Version History

Introduced in R2024b

See Also

| |