Re: init_htab causes SIGFPE (or worse) due to miscalculation for large nbuckets

On Mon, 2012-12-10 at 21:19 -0500, Tom Lane wrote:
> I hadn't gone any further than to code and test the functions I listed.
> If you are working on a complete patch, please press on.

Attached. This fixes my test case[1] and uses the functions that you
wrote. I made them static because I couldn't think of a reason for
something outside to call them.

Regards,
    Jeff Davis

[1]: The test case will just eat a lot of memory right now, but that's
only because I set work_mem so high. So, it doesn't actually complete,
but it no longer corrupts the HASHHDR.