Determining fastest must be a matter of profiling and benchmarking. And ultimately if speed is your greatest concern you would drop into assembly language and hand-craft your routines. If speed isn't quite that critical, here is one way to do it:

my %hash;
while ( my $line = <DATA> ) {
    print "Found another duplicate of: \"",
          $line,
          "\" in line $.\n" 
        if ++$hash{$line} > 1;
}
[download]

That will tell you if the file contains duplicate lines.

On the other hand, if you have a particular line in mind, and want to know if it is repeated elsewhere, you pretty much just scan through the file looking for a duplicate. No shortcut there.