From: | tgl(at)postgresql(dot)org (Tom Lane) |
---|---|
To: | pgsql-committers(at)postgresql(dot)org |
Subject: | pgsql: Fix best_inner_indexscan to return both the cheapest-total-cost |
Date: | 2007-05-22 01:40:33 |
Message-ID: | 20070522014033.C60D19FB293@postgresql.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-committers |
Log Message:
-----------
Fix best_inner_indexscan to return both the cheapest-total-cost and
cheapest-startup-cost innerjoin indexscans, and make joinpath.c consider
both of these (when different) as the inside of a nestloop join. The
original design was based on the assumption that indexscan paths always
have negligible startup cost, and so total cost is the only important
figure of merit; an assumption that's obviously broken by bitmap
indexscans. This oversight could lead to choosing poor plans in cases
where fast-start behavior is more important than total cost, such as
LIMIT and IN queries. 8.1-vintage brain fade exposed by an example from
Chuck D.
Modified Files:
--------------
pgsql/src/backend/nodes:
outfuncs.c (r1.306 -> r1.307)
(http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/nodes/outfuncs.c.diff?r1=1.306&r2=1.307)
pgsql/src/backend/optimizer/path:
indxpath.c (r1.221 -> r1.222)
(http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/optimizer/path/indxpath.c.diff?r1=1.221&r2=1.222)
joinpath.c (r1.111 -> r1.112)
(http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/optimizer/path/joinpath.c.diff?r1=1.111&r2=1.112)
pgsql/src/include/nodes:
relation.h (r1.141 -> r1.142)
(http://developer.postgresql.org/cvsweb.cgi/pgsql/src/include/nodes/relation.h.diff?r1=1.141&r2=1.142)
pgsql/src/include/optimizer:
paths.h (r1.97 -> r1.98)
(http://developer.postgresql.org/cvsweb.cgi/pgsql/src/include/optimizer/paths.h.diff?r1=1.97&r2=1.98)
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2007-05-22 01:40:42 | pgsql: Fix best_inner_indexscan to return both the cheapest-total-cost |
Previous Message | Tom Lane | 2007-05-21 17:57:35 | pgsql: Teach tuplestore.c to throw away data before the "mark" point |