Here data structure is the key. It looks like you are using three arrays in parallel to manage data records. Instead you should be using an array of records - each record being either a hash or an array as you think most appropriate. Consider:

#!/usr/bin/perl
use strict;
use warnings;
use Text::CSV;

my @ports = qw(portname portID port802);
my @val1  = qw(ON ON OFF);
my @val2  = qw(EX EX NEX);

my @records;

while (@ports) {
    push @records, [shift @ports, shift @val1, shift @val2];
}

my $csv = Text::CSV->new ({sep_char => ';'});

for my $record (@records) {
    if ($csv->combine (@$record)) {
        print $csv->string, "\n";
    } else {
        print "combine () failed on argument: ", $csv->error_input, "\
+n";
    }
}
[download]

Prints:

portname;ON;EX
portID;ON;EX
port802;OFF;NEX
[download]

Note that much of the rest of your program would likely benefit from using record oriented storage as illustrated rather than scattering related data across disparate arrays as indicated in your sample.

Or, much more consice:

#!/usr/bin/perl
use strict;
use warnings;
use Text::CSV;

my @ports = qw(portname portID port802);
my @val1  = qw(ON ON OFF);
my @val2  = qw(EX EX NEX);

my $csv = Text::CSV->new ({sep_char => ';', eol => "\n", auto_diag => 
+1 });

for (@ports) {
    $csv->print (*STDOUT, [ $_, shift @val1, shift @val2 ]);
    }
[download]


Enjoy, Have FUN! H.Merijn

Concise was not the point. Adopting an appropriate data structure was. Note that the meat of my reply however boils down to:

$csv->print (*STDOUT, $_) for @records;
[download]

where your 'concise' equivalent was:

for (@ports) { 
    $csv->print (*STDOUT, [ $_, shift @val1, shift @val2 ]); 
}
[download]

which after all is part of why an appropriate data structure is a 'good thing'™ ;-)

True laziness is hard work

sub combine_rows
{
    my ( $label, @arrays ) = @_;

    my @joined;
    for my $i ( 0 .. $#{$arrays[0]} ) {  # assumes that all arrays are
+ the same size
        push @joined, join( ';', $label, map { $arrays[$_][$i] } 0 .. 
+$#arrays );
    }
    return \@joined;
}
[download]

@ports = qw( portname portID port802 );
@val1 = qw( ON ON OFF );
@val2 = qw( EX EX NEX );

my $joined = combine_rows( "ports", \@ports, \@val1, \@val2 );

print "$_\n" for ( @$joined );
[download]

The reason I use "map" in the subroutine is to allow it to join any number of parallel arrays (not just exactly 3, as in your "@ports, @val1, @val2" example). If you can't figure out what "map" is doing there, it's just a nested loop to get the $i'th value of each input array.