Check integer conversion for overflow in datetime functions. Problem
reported by Christopher Kings-Lynne:
http://archives.postgresql.org/pgsql-hackers/2005-11/msg01385.php
In one place (line 60 in the patch) the code sets errno to 0 when
it should still be 0 after similar code a few lines above (otherwise
the function would have returned an error). I wasn't sure what
would be preferred: clearing errno again "just to be sure," a comment
explaining why it isn't necessary, or nothing at all.
This patch should apply cleanly against HEAD and 8.1. If the patch
looks good then I'll submit patches for earlier branches when I get
a chance to build and test those versions.
--
Michael Fuhr