Here is MCE::Flow where workers read from the input file directly. MCE->gather is called to append to an array.

#!/usr/bin/env perl

use strict;
use warnings;

use MCE::Flow;
use IO::Socket;

my $nthreads = 20;
my $in_file2 = 'rang.txt';

sub ip_checker {
    my $ip = $_;
    chomp($ip);
    my $host = IO::Socket::INET->new(
        PeerAddr => $ip,
        PeerPort => 80,
        proto    => 'tcp',
        Timeout  => 1
    );
    if ( defined $host ) {
        MCE->gather($ip);
    }
}

my @result = MCE::Flow->run_file(
    { chunk_size => 1, max_workers => $nthreads },
    \&ip_checker,
    $in_file2,
);

MCE::Flow->finish();

open( my $output, ">", "port.txt" ) or die $!;
print {$output} "$_\n" for @result;
close($output);
[download]

For this one, workers write serially to the output file via MCE->print.

use strict;
use warnings;

use MCE::Flow;
use IO::Socket;

my $nthreads = 20;
my $in_file2 = 'rang.txt';
open( my $output, ">", "port.txt" ) or die $!;

sub ip_checker {
    my $ip = $_;
    chomp($ip);
    my $host = IO::Socket::INET->new(
        PeerAddr => $ip,
        PeerPort => 80,
        proto    => 'tcp',
        Timeout  => 1
    );
    if ( defined $host ) {
        MCE->print($output, "$ip\n");
    }
}

MCE::Flow->run_file(
    { chunk_size => 1, max_workers => $nthreads },
    \&ip_checker,
    $in_file2,
);

MCE::Flow->finish();
close($output);
[download]

Perl provides various ways to concur parallel processing. There is Parallel::ForkManager. For threads-like API, there is MCE::Child and MCE::Hobo. MCE itself provides chunking capabilities. It has sugar syntax to gather into an array or hash. For output, MCE->print, MCE->printf, and MCE->say write serially to a file handle (default STDOUT if omitted).

Basically, no problem if the Perl binary lacks threads support.