Re: Plan time Improvement - 64bit bitmapset - Mailing list pgsql-hackers

From Andres Freund
Subject Re: Plan time Improvement - 64bit bitmapset
Date
Msg-id 4A26A613.70102@anarazel.de
Whole thread Raw
In response to Re: Plan time Improvement - 64bit bitmapset  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: Plan time Improvement - 64bit bitmapset  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Hi,

On 06/03/2009 06:21 PM, Tom Lane wrote:
> Andres Freund<andres@anarazel.de>  writes:
>> When switching bitmapword and companions in bitmap.h to u64 and s64
>> respectively I get an improvement up to 15% in queries with 16+ joins.
> I find this *really* hard to believe, because I've never seen the bitmap
> support operations show up noticeably at all in profiles.  What sort of
> queries are you testing?
Many left joins from one base relation to additional dimensions. All the 
dimensions are relatively complex views consisting out of multiple joins 
or subselects.
Some correlated subqueries and some [NOT] EXISTS() are also included in 
some of the queries.


I tested by compiling with 64bit bitmaps and without by repeatedly just 
changing those three definitions. I don't see how I could get false 
results with that?

I guess the biggest advantage comes from less cache trashing?

Andres


pgsql-hackers by date:

Previous
From: Andres Freund
Date:
Subject: Re: Managing multiple branches in git
Next
From: Andrew Dunstan
Date:
Subject: Re: Managing multiple branches in git