Video Demo Using VICP on DM648

Do you run any algorithm on DSP except the VICP library ?

If not, then very likely the 100% load is due to the way you synchronizing with VICP interrupt: the example shipped with the library uses busy waiting for a flag inside CPIS_wait():

void CPIS_isr() {

    VICPIntFlag= 1;

}

Int32 CPIS_wait_Fxn(void *arg) {

    while (VICPIntFlag==0);

        VICPIntFlag= 0;

    return 0;

}

However this is not efficient as the while (VICPIntFlag==0); actively uses the CPU to poll for the VICPIntFlag.  The most efficient way is to use semaphore instead of global flag VICPIntFlag.

void CPIS_isr() {

   SEM_post(vicp_semaphore);

}

Int32 CPIS_wait_Fxn(void *arg) {

    SEM_pend(vicp_semaphore);

    return 0;

}

The global variable vicp_sempahore of type SEM_handle will have to be created in main() using SEM_create().

regards,

Victor