running error

[ceceNo1]

When I ran this program, it was stuck in a step and couldn't go forward.
optimiz.c:
1601 alam = alam2 = f2 = fold2 = tmplam = .0;
1675 a=(rhs1/(alamalam)-rhs2/(alam2alam2))/(alam-alam2);
1676 b=(-alam2rhs1/(alamalam)+alamrhs2/(alam2alam2))/(alam-alam2);
While debugging and analyzing the source code, I noticed that the initial value of double alam2 in optimiz.c:Lnsrch was 0, and alam2 was not assigned again in the subsequent calculations. Thus in the calculation of optimiz.c:1675-1676, alam2 appears as the denominator. This results in a division by 0 situation, where a and b are both Nan. I would like to ask the author if this is a tacit (design thought) situation or if it is triggered by my usage.