TMS320C5535: DSPlib mul32() not behaving as expected

I'm developing on a TMS320C5535 using the XDS110 debugger, the C5000 chip support library, and Code Composer version: 9.1.0.00010. I'm having a hard time getting mul32() to work in version 03.00.00.03 of DSPlib. To demonstrate where I'm at here is an example of me using add() of DSPlib:

int main(void)
{
    /////////////* Initializations */////////////
    CSL_Status SYS_status;

    SYS_status = SYS_PLL_LDO_init();
    SYS_status = SYS_EBSR_init();
    SYS_status = SYS_GPIO_init();

#define NX 10
    Uint32 i;
//    LDATA x[NX] = {0};
//    LDATA y[NX] = {0};
//    LDATA z[NX] = {0};
    DATA x[NX] = {0};
    DATA y[NX] = {0};
    DATA z[NX] = {0};

    for (i = 0; i < NX; i++){
        x[i] = i;
        y[i] = i;
    }

    add(x, y, z, NX, 0);
//    mul32(x, y, z, NX);

    /////////////* Infinite Loop */////////////
    while(1){
        i = 0;
    }

}

x and y are both =  [0,1,2,3, ... ,9]. As expected, z = [0, 2, 4, 6, ... ,18]. Now I try using mul32():

/////////////* Main Function */////////////
int main(void)
{
    /////////////* Initializations */////////////
    CSL_Status SYS_status;

    SYS_status = SYS_PLL_LDO_init();
    SYS_status = SYS_EBSR_init();
    SYS_status = SYS_GPIO_init();

#define NX 10
    Uint32 i;
    LDATA x[NX] = {0};
    LDATA y[NX] = {0};
    LDATA z[NX] = {0};
//    DATA x[NX] = {0};
//    DATA y[NX] = {0};
//    DATA z[NX] = {0};

    for (i = 0; i < NX; i++){
        x[i] = i;
        y[i] = i;
    }

//    add(x, y, z, NX, 0);
    mul32(x, y, z, NX);

    /////////////* Infinite Loop */////////////
    while(1){
        i = 0;
    }

}

x and y are still  both =  [0,1,2,3, ... ,9]. However, z is incorrectly = [0,0, .... 0].

SPRU422J mentions that "Input and Output vectors must be 32−bit aligned", but I do now know how to do this or if it's the reason this unexpected behavior is occurring. Any guidance would be greatly appreciated