Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:85900
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmail.com designates 74.125.82.45 as permitted sender)
MIME-Version: 1.0
Sender: julienpauli@gmail.com
In-Reply-To: <55342A50.9030100@gmail.com>
References: <CAMUwpuTZK=zqHcWG21DA5fUSODuq6njyx9+RUtQR0gQu02a5AQ@mail.gmail.com>
 <55342A50.9030100@gmail.com>
Date: Wed, 22 Apr 2015 14:18:21 +0200
Message-ID: <CAMUwpuQUFa0G5wB=Bz_KJEDVOkCpaAEAfonHgQEh_MnzwhA-sA@mail.gmail.com>
To: Stanislav Malyshev <smalyshev@gmail.com>
Cc: Andrew Faulds <ajf@ajf.me>, PHP Internals <internals@lists.php.net>
Content-Type: multipart/alternative; boundary=089e0122ec88188fe805144f2b7d
Subject: Re: [PHP-DEV] Closure::call() to access private data, really ?
From: jpauli@php.net (Julien Pauli)

--089e0122ec88188fe805144f2b7d
Content-Type: text/plain; charset=UTF-8

On Mon, Apr 20, 2015 at 12:21 AM, Stanislav Malyshev <smalyshev@gmail.com>
wrote:

> Hi!
>
> > Just a simple use case showing how dangerous that is :
> >
> > <?php
> > $p = function($p) { $this->$p = new Stdclass; };
> > $p->call($e = new Exception, 'trace');
> > throw $e;
>
> Yes, this is not good, and this is the consequence of allowing to rebind
> closures. I'm not sure though how to fix it except for banning closures
> from assuming scope of internal classes. If you assigned the scope of
> Exception to it, it should have access to Exception - that's how the
> scope works.
>
> > For User classes, this could lead to information leaks or bad behaviors,
>
> I'm not sure what you mean by "information leaks", but the behavior is
> on whoever wrote that code. People can write bad code, we can't disallow
> this. We can restrict things that would really break (like segfault) but
> I'm not sure what else we can do here. If you have the scope of the
> class, that implies access to the private members of the class. So we
> either have to not let the closure have the scope, or invent some
> additional term of scope' that is not like real scope. I don't think
> it'd be good.
>


Yup Stas.

Perhaps the best thing to do is to forbid rebinding a Closure to an
internal class ?

I'm sure we could find segfaulting behaviors using such a trick on most of
our internals classes, aka mysqli, simplexmlelement, PDO or SPL classes.

Thoughts ?

Julien.P

--089e0122ec88188fe805144f2b7d--