From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <e@yhbt.net>
X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on dcvr.yhbt.net
X-Spam-Level: 
X-Spam-Status: No, score=-4.0 required=3.0 tests=ALL_TRUSTED,BAYES_00
	shortcircuit=no autolearn=ham autolearn_force=no version=3.4.2
Received: from localhost (dcvr.yhbt.net [127.0.0.1])
	by dcvr.yhbt.net (Postfix) with ESMTP id A05701F55B;
	Thu, 11 Jun 2020 19:39:21 +0000 (UTC)
Date: Thu, 11 Jun 2020 19:39:21 +0000
From: Eric Wong <e@yhbt.net>
To: meta@public-inbox.org
Subject: amusing CoW string dedupe example
Message-ID: <20200611193921.GA17563@dcvr>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
List-Id: <meta.public-inbox.org>

I've always known hash keys get deduplicated in Perl to save RAM
for a while; but it turns out that it's possible to (ab)use them
for dynamic strings even when the hash doesn't live beyond a
subroutine scope.

Looks like we'll be able to save RAM in some places :D

The following script takes between 5-9M of RAM depending on the
version of Perl I'm using (tested down to 5.16.3), but nearly 1G
if the early return is uncommented in the `dedupe' sub:

----8<----
#!perl -w
use strict;
use Devel::Peek;
sub dedupe {
	my ($k) = @_;
#return $k; # uncomment to disable CoW dedupe, needs ~1G of RAM
	my %cow = ( $k => undef );
	((keys %cow)[0]);
}

my $n = 3_600_000;
my $k = pack('S*',($n + 1)..($n + 50000)); # uint16_t array[50000]
my @x = map { dedupe($k) } (1..10000);

{
	no warnings 'once';
	$Devel::Peek::pv_limit = 10;
}
Dump(\@x);
print 'length: ',length($k), "\n";