This depends on the paper and the conference! I seen the crash 2 paper presented after it was published! If the paper is significant enough and the conference agrees that it is worthy despite publication then why not? Its worth liaising with the organisation that one has published the paper and two the conference to confirm both are happy! But generally as a rule i would say not good practice!
So far whichever conference i have attended they did not mentioned any rule regarding your question. Well in my opinion a good research should always reach its audience so we are communicating our research to the right audience, so you should present your paper in the conference whether it is published or not, with slight modifications.
Firstly, I appreciate your concern. Coming to the point, there are no certain rules & regulations for this (i.e. your question). However, it is generally said that we are not supposed to do that. (viz., which is not general practice). And it also depends on what level conference you wanted to present (either local/national/international level).
It is just an assumption or generalized self-framed perception, you have all the right to re-frame it. (just my personal opinion) However, as Jonathan Herron said if that particular paper is unique and novel in nature OR if you think there is an immense need for that particular topic in present scenario and you want public (or certain population) to get to know regarding this, then definitely YES, you can go for it in order to enhance the visibility of your study.
As most of the conferences never frame any sort of such rules, I assume you can do this, unless it is an international conferences where the abstracts/full text articles of the conference would precede for publication.
But, my personal suggestion would be, if the study was conducted long back OR if its been few yeas since it got published OR if you had already presented it in ample number of conferences (same paper) then DO NOT follow this as its already been reached its saturation stage. However, I recommend you to have a glance at a paper recommended by Manik Chhabra in order to get better idea of it.