My approach: I have implemented a segment tree for finding max in
[l,r]. Whenever any query is asked I find out the maximum element in
[l,r] and then update the segment tree by replacing the maximum element found by
-INF.Now, I find out the maximum element which gives me the second largest element in
[l,r].At last, I update the first largest back to its initial value.
I am getting TLE. How can I optimize the updation part?