First of all, the question is whether x or y can be assumed to have unique values. Either of the two can then be used as keys to a hash to store the association of the two arrays. In this case, the task is straightforward.

In the general case, one needs to introduce the position as the key to a hash that stores x and y. One can then sort the keys with respect to y. Here is how it could look like:

use strict;
use warnings;
# initial data
my @x   = qw/c d e f k l m n/; 
my @y   = qw/4 6 5 2 9 7 8 3/; 
# store in hash to keep association between x and y after sorting
my %data;
for my $i (0..@x-1) {
        $data{$i}{x} = $x[$i]; # store x corresponding to position $i
        $data{$i}{y} = $y[$i]; # store y corresponding to position $i
}
# sort positions with respect to y
my @sorted = sort { $data{$a}{y} <=> $data{$b}{y} } keys %data;
print "Sorted data:\n";
for my $i (@sorted) {
        print "$data{$i}{y} corresponds to $data{$i}{x} at position $i
+\n";
}
[download]

To print the highest and lowest 5 should now be simple...