On Fri, Mar 02, 2012 at 03:45:38PM -0500, Robert Haas wrote:
> SELECT SUM(1) FROM (SELECT * FROM randomtext ORDER BY t) x;
>
> On unpatched master, this takes about 416 ms (plus or minus a few).
> With the attached patch, it takes about 389 ms (plus or minus a very
> few), a speedup of about 7%.
>
> I repeated the experiment using the C locale, like this:
>
> SELECT SUM(1) FROM (SELECT * FROM randomtext ORDER BY t COLLATE "C") x;
>
> Here, it takes about 202 ms with the patch, and about 231 ms on
> unpatched master, a savings of about 13%.
> [oprofile report, further discussion]
Thanks for looking into this. Your patch is also a nice demonstration of
sortsupport's ability to help with more than just fmgr overhead.
> Considering all that, I
> had hoped for more like a 15-20% gain from this approach, but it
> didn't happen, I suppose because some of the instructions saved just
> resulted in more processor stalls. All the same, I'm inclined to
> think it's still worth doing.
This is a border case, but I suggest that a 13% speedup on a narrowly-tailored
benchmark, degrading to 7% in common configurations, is too meager to justify
adopting this patch.
nm
