[keszler@tek perl]# cat test3.pl
#!/usr/bin/perl -w

use Benchmark;
use strict;

my $results = timethese( 1e4, {
  a => sub{print "\n" for(1..1e4);},
  b => sub{print $/ for(1..1e4);},
} );

my $results2 = timethese( 1e4, {
  a2 => sub{print $/ for(1..1e4);},
  b2 => sub{print "\n" for(1..1e4);},
} );
[keszler@tek perl]# ./test3.pl | grep -v '^$'
Benchmark: timing 10000 iterations of a, b...
         a: 54 wallclock secs (33.83 usr +  0.92 sys = 34.75 CPU) @ 287.77/s (n=10000)
         b: 54 wallclock secs (34.21 usr +  0.96 sys = 35.17 CPU) @ 284.33/s (n=10000)
Benchmark: timing 10000 iterations of a2, b2...
        a2: 54 wallclock secs (34.37 usr +  0.93 sys = 35.30 CPU) @ 283.29/s (n=10000)
        b2: 54 wallclock secs (33.69 usr +  0.90 sys = 34.59 CPU) @ 289.10/s (n=10000)