You can certainly do alot better than that long piece of code for this task though. The last two loops can be combined:

for (sort { $module{$a} <=> $module{$b} } keys %module) {
    printf "%04d, %s\n", $module{$_}, $_;
}
[download]

Gets you rid of the synthetic @list too. I have no idea what your if/elsif are supposed to sort out, either. And finally, instead of using <DATA>, you can use <> and just pipe the output from grep to the script. A compressed version:

/^use\s+([^\s;]+)/ && $_{$1}++ while <>;

printf "%04d, $_\n", $_{$_} for sort { $_{$b} <=> $_{$a} } keys %_;
[download]

And then it's just

$ grep -hr ^use perl/ | perl foo.pl

Makeshifts last the longest.