What is the time complexity of the following dynamic programming implementation of the maximum sum rectangle problem?

int max_sum_rectangle(int arr[][3],int row,int col)
{
      int left, right, tmp[row], mx_sm = INT_MIN, idx, val;
      for(left = 0; left < col; left++)
      {
           for(right = left; right < col; right++)
           {
               if(right == left)
               {
                   for(idx = 0; idx < row; idx++)
                     tmp[idx] = arr[idx][right];
               }
               else
               {
                   for(idx = 0; idx < row; idx++)
                      tmp[idx] += arr[idx][right];
               }
               val = kadane_algo(tmp,row);
               if(val > mx_sm)
                 mx_sm = val;
           }
      }
      return mx_sm;
}

A. O(row*col)

B. O(row)

C. O(col)

D. O(row*col*col)

Answer: Option D