ixid:
int iter_next = iter + 1 > k? 0 : iter + 1;
This is becoming a "fixed size circular queue". But maybe a modulus is faster than a branch here. (It's better when k is always a power of two, you don't need a modulus. And even better if your size is a multiple of the page size).
nums[iter_next] = total % 10^^8;
In such cases I suggest to add (), to not force the person that reads the code reader to remember the precedence of uncommon operators.
Bye, bearophile
