"Thomas F. O'Connell" <tfo@sitening.com> writes:
> I've brought this up to an extent in the past, but is there an easy
> way to extend section 12.3.1 (or create some form of appendix) such
> that it reveals all possible locking paths for SQL commands in
> postgres?
It's very simple: any form of schema alteration on a table takes
exclusive lock. Offhand the only exception I can remember is CREATE
INDEX, and frankly I'm not 100% sure that is a safe exception.
regards, tom lane