You can retrieve the code from named subs or coderefs with B::Deparse:

use B::Deparse;
my $d = B::Deparse->new;
# anonymous sub
my $coderef = sub { print "howdy!" }; # assuming herveus ;)
my $code = $d->coderef2text($coderef);
# named sub, e.g. coderef2text from B::Deparse
$code = $d->coderef2text(\&B::Deparse::coderef2text);
[download]

--shmem

_($_=" "x(1<<5)."?\n".q·/)Oo.  G°\        /
                              /\_¯/(q    /
----------------------------  \__(m.====·.(_("always off the crowd"))."·
");sub _{s./.($e="'Itrs `mnsgdq Gdbj O`qkdq")=~y/"-y/#-z/;$e.e && print}

CODE REFERENCES
       Since Storable version 2.05, CODE references may be seri-
       alized with the help of B::Deparse. To enable this fea-
       ture, set $Storable::Deparse to a true value. To enable
       deserializazion, $Storable::Eval should be set to a true
       value. Be aware that deserialization is done through
       "eval", which is dangerous if the Storable file contains
       malicious data. You can set $Storable::Eval to a subrou-
       tine reference which would be used instead of "eval". See
       below for an example using a Safe compartment for deseri-
       alization of CODE references.

       If $Storable::Deparse and/or $Storable::Eval are set to
       false values, then the value of $Storable::forgive_me (see
       below) is respected while serializing and deserializing.
[download]

I'm curious about your application. Do you want your client to be able to run any arbitrary piece of Perl code? Is this wise? Apart from the security risks if your client is compromised, how would you intend to debug code running on the server?

Granted you could run in safe mode, and restrict anything dangerous, but a better approach might be straightforward RPC; any code you need to run would already exist server side, you merely do method calls to it via a proxy object. This would make developing and testing it a whole site easier.