ARM CMSIS-DSP

lbuild module: modm:cmsis:dsp

This module provides the CMSIS-DSP library. Please see the API documentation for details.

Note: - The library should be accessed through the arm_math.h and arm_math_f16.h headers, as these include necessary configuration options. - This module splits up the library into submodules to speed up compilation by only including the math functions you really need. - You can enable 16-bit floating point support via the option with_f16.

Configuration

You can define your own configuration in the <arm_math_local.h> file which will automatically be included at the beginning of the arm_math.h and arm_math_f16.h files. This way changes to your CMSIS-DSP configuration will only affect its source files and not recompile your whole project.

Example <arm_math_local.h> configuration:

// Check the input and output sizes of matrices
#define ARM_MATH_MATRIX_CHECK 1
// enable manual loop unrolling in DSP functions
#define ARM_MATH_LOOPUNROLL 1

Options

with_f16

Include f16 format functions

16-bit floating point is only used for storage, since the Cortex-M FPU does not support the format in hardware.

Default: no
Inputs: [yes, no]

Limited availability: Check with 'lbuild discover' if this module is available for your target!