I think this is what you want:

# cat 791596.pl
use strict;
use warnings;
use Data::Dumper;

my %h1 = ( e=>'f', c=>'d', a=>'b' );
my %h2 = ( e=>'K', a=>'J', q=>'V' );

do { 
   my @b = grep {exists $h1{$_}} keys %h2;
   @h1{@b} = @h2{@b};
};

print Dumper(\%h1);
# perl 791596.pl
$VAR1 = {
          'e' => 'K',
          'c' => 'd',
          'a' => 'J'
        };
[download]

The first hash is updated with the values from the second, but no new keys (such as q in the second hash) are added to the first hash.

Update: Correction as per Gramps' reply. ;^)