Our arm_math.h (CMSIS-DSP) library is too out of date, that some known issues become growing concerns.
For example, for our legacy version of library, arm_sin_f32 have erroneous output when input is within the range (-10^(-6.727),0), and we'd have to use AHRS_sinf as a temporary fix. Check https://www.keil.com/pack/doc/CMSIS_Dev/DSP/html/ChangeLog_pg.html for reference, the related change log is "corrected arm_sin_f32() for very small negative input values".
Our arm_math.h (CMSIS-DSP) library is too out of date, that some known issues become growing concerns.
For example, for our legacy version of library, arm_sin_f32 have erroneous output when input is within the range (-10^(-6.727),0), and we'd have to use AHRS_sinf as a temporary fix. Check https://www.keil.com/pack/doc/CMSIS_Dev/DSP/html/ChangeLog_pg.html for reference, the related change log is "corrected arm_sin_f32() for very small negative input values".