Horton

Re: reducing the footprint of ScanKeyword (was Re: Large writable variables)

Alvaro Herrera <alvherre@2ndquadrant.com>

View thread View on postgr.es
From: Alvaro Herrera <alvherre@2ndquadrant.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Andrew Dunstan <andrew.dunstan@2ndquadrant.com>, John Naylor <jcnaylor@gmail.com>, Robert Haas <robertmhaas@gmail.com>, Andres Freund <andres@anarazel.de>, Andrew Gierth <andrew@tao11.riddles.org.uk>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2018-12-27T22:04:41Z
Lists: pgsql-hackers 
On 2018-Dec-27, Tom Lane wrote:

> I poked around a little on my own machines, and I can confirm that
> Getopt::Long is present in a default Perl install-from-source at
> least as far back as perl 5.6.1.  It's barely conceivable that some
> packager might omit it from their minimal package, but Red Hat,
> Apple, NetBSD, and OpenBSD all include it.  So it sure looks to
> me like relying on it should be non-problematic.

In Debian it's included in package perl-modules-5.24, which packages
perl and libperl5.24 depend on.  I suppose it's possible to install
perl-base and not install perl-modules, but it'd be a really bare-bones
machine.  I'm not sure it's possible to build Postgres in such a
machine.

-- 
Álvaro Herrera                https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

Commits

  1. Use perfect hashing, instead of binary search, for keyword lookup.

  2. Reduce the size of the fmgr_builtin_oid_index[] array.

  3. Replace the data structure used for keyword lookup.