Sort the letters, sort the letters of the words of a dictionary file, and match them up.

sub canonize { return join '', sort split //, lc $_[0]; }

my %dict;
{
   open(my $fh, '<', $dict_file)
      or die("Unable to open dictionary file \"$dict_file\": $!\n");

   while (<$fh>) {
      chomp;
      push @{ $dict{canonize($_)} }, $_;
   }
}

my $jumble = 'OLHOSC';

$jumble = canonize($jumble);
if (exists($dict{$jumble})) {
   print("$_\n") for @{ $dict{$jumble} };
} else {
   print("No matches found.\n");
}
[download]

You might want to look at using a Trie for a more memory efficient structure than a hash for this purpose.

If you only want one match, and you're only working on one word, you can just process the dictionary file until you find a match.

sub canonize { return join '', sort split //, lc $_[0]; }

my $jumble = 'OLHOSC';

open(my $fh, '<', $dict_file)
   or die("Unable to open dictionary file \"$dict_file\": $!\n");

my $found = 0;
$jumble = canonize($jumble);
while (<$fh>) {
   chomp;
   if (canonize($_) eq $jumble) {
      print("$_\n");
      $found = 1;
      last;
   }
}

if (!$found) {
   print("No matches found.\n");
}
[download]

Update: The others inlined canonize. That's probably a good idea since function calls are expensive and you'll be canonizing a lot of words.