It’s not much, but it’s not nothing, my first ever patch to PostgreSQL proper is part of yesterday’s 8.4.2 release:

Fix incorrect logic for GiST index page splits, when the split depends on a non-first column of the index (Paul Ramsey)

It was just a minor syntax error I found because I was reading through that section of the code very, very, very closely (it’s true, I copy the work of smarter people than I) while implementing the indexes for GEOGRAPHY in PostGIS 1.5.