Re: [HACKERS] Performance improvement for joins where outer side is unique

On 27 January 2017 at 08:34, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> David Rowley <david.rowley@2ndquadrant.com> writes:
>> I've attached a version without outer unique.
>
> I looked through this a bit, and the first thing I noticed was it doesn't
> touch costsize.c at all.  That seems pretty wrong; it's little help to
> have a performance improvement if the planner won't pick the right plan
> type.  There were costsize.c changes in there back in April; what happened
> to them?

hmm. I must've taken them out when I changed everything around to use
the join types. Nothing special was needed when changing JOIN_INNER to
JOIN_SEMI.

I must've forgotten to put them back again... I'll do that now.

-- David Rowley                   http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training & Services