We have a couple of unfixed bugs regarding extensions, for which patches
have been proposed but remain unfixed in git.
The oldest is bug #6704, for which a proposed fix for the master branch
was posted here:
http://archives.postgresql.org/message-id/m2zk4e6p7m.fsf@2ndQuadrant.fr
There was some disagreement about the proper way forward, so I hadn't
looked at this patch. I just did, and find that even if the approach
taken by the patch is the correct one, it needs some polish before it
can be committed.
The other one was reported by Marko Kreen in
http://archives.postgresql.org/message-id/CACMqXCJjauc9jPa64VxskRN67bYjuYmodz-Mgy-_aoZ6ErG3Yg@mail.gmail.com
(the thread talks about 2 bugs, but one of them is just pilot error).
Dimitri posted patches to the real bug for the three affected branches;
the last one (for master) was in message
http://archives.postgresql.org/message-id/m2391yy6ub.fsf@2ndQuadrant.fr
I eyeballed this patch and it seemed to me to fix the problem, but I am
unfamiliar enough with this dark corner of pg_dump that I can't be sure
that there aren't bad behaviors being introduced inadvertently.
--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services