CCS/TMS320F28377D: Using functions in standard math library throws ILLEGAL_ISR

Part Number: TMS320F28377D

Tool/software: Code Composer Studio

Hi,

I want to make a time-analysis for math operation between CPU and CLA. I want to use exponential function for this analysis.

I can successfully run exponantial function on CLA, it takes around 67 SYSCLK cycles. However when I try to use exp function in standard maht libray on CPU (not CLA) it throws ILLEGAL_ISR.

The line where this exception occurs in file e_expf.c on line 101. I copy whole function below.

My code is simple: 

float NumDen;
double temp;

NumDen = -1.56f;

temp = exp(NumDen); // Exception occurs

float expf(float x) function in e_expf.c

float
expf(float x)
{
	float y,hi=0.0,lo=0.0,c,t,twopk;
	int32_t k=0,xsb;
	uint32_t hx;

	GET_FLOAT_WORD(hx,x);
	xsb = (hx>>31)&1;		/* sign bit of x */
	hx &= 0x7fffffff;		/* high word of |x| */

    /* filter out non-finite argument */
	if(hx >= 0x42b17218) {			/* if |x|>=88.721... */
	    if(hx>0x7f800000)
		 return x+x;	 		/* NaN */
            if(hx==0x7f800000)
		return (xsb==0)? x:0.0;		/* exp(+-inf)={inf,0} */
	    if(x > o_threshold)
            {
                __raise_overflow();
                return INFINITY;
            }
	    if(x < u_threshold)
            {
                __raise_underflow();
                return 0;
            }
	}

    /* argument reduction */
	if(hx > 0x3eb17218) {		/* if  |x| > 0.5 ln2 */
	    if(hx < 0x3F851592) {	/* and |x| < 1.5 ln2 */
		hi = x-ln2HI[xsb]; lo=ln2LO[xsb]; k = 1-xsb-xsb;
	    } else {
		k  = invln2*x+halF[xsb];
		t  = k;
		hi = x - t*ln2HI[0];	/* t*ln2HI is exact here */
		lo = t*ln2LO[0];
	    }
	    STRICT_ASSIGN(float, x, hi - lo);
	}
	else if(hx < 0x39000000)  {	/* when |x|<2**-14 */
	    if(huge+x>one) return one+x;/* trigger inexact */
	}
	else k = 0;

    /* x is now in primary range */
	t  = x*x;
	if(k >= -125)
	    SET_FLOAT_WORD(twopk,0x3f800000+(k<<23));
	else
	    SET_FLOAT_WORD(twopk,0x3f800000+((k+100)<<23));
	c  = x - t*(P1+t*P2);
	if(k==0) 	return one-((x*c)/(c-(float)2.0)-x);
	else 		y = one-((lo-(x*c)/((float)2.0-c))-hi);
	if(k >= -125) {
	    if(k==128) return y*2.0F*0x1p127F; // LINE 101: ERROR OCCURS HERE
	    return y*twopk;
	} else {
	    return y*twopk*twom100;
	}
}

I suspect some overflow? How can I run it succesfully?

Thanks and merry christmas.