Changing authorship order without consent is not common in academia, and this should not have happened. However, rather than asking yourself “what can you do”, as @Dirk Liebhold pointed out, I would ask “what do I want”. Are you going to keep on working with your advisor for a long time or are you leaving immediately after? Is he/she reasonable? If you do not plan on publishing with him, do you really care about pointing out his misconduct?
Keep in mind that last author is not so bad: first author carried out the experiment, but last author is the one who designed or had the idea for it, and possibly the one who found the money to fund it.
In my opinion it is not very important if you are first or last
, given that this is only a conference paper , but you should make it clear who is actually going to the conference! Also, if a paper is going to follow on this work, decide the author list in advance with him.
Edit: In some fields (such as computer science), conference papers are peer-reviewed and are at least as important as journal articles, and last authorship carries no prestige.