Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:36168
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain zend.com designates 212.25.124.162 as permitted sender)
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Mon, 17 Mar 2008 14:23:54 +0200
Message-ID: <06B0D32C7A96544490D18AF653D6BDE502759A5A@il-ex1.zend.net>
Thread-Topic: [PHP-DEV] Patch for opcode caches
Thread-Index: AciGAdz0ykjod3GsQJWDcZ4ZsLpiXACJdBGQ
To: "Marcus Boerger" <mail@marcus-boerger.de>,
	"Stas Malyshev" <stas@zend.com>
Cc: <internals@lists.php.net>
Subject: RE: [PHP-DEV] Patch for opcode caches
From: dmitry@zend.com ("Dmitry Stogov")



> -----Original Message-----
> From: Marcus Boerger [mailto:mail@marcus-boerger.de]=20
> Sent: Friday, March 14, 2008 9:28 PM
> To: Stas Malyshev
> Cc: internals@lists.php.net
> Subject: Re: [PHP-DEV] Patch for opcode caches
>=20
>=20
> Hello Stanislav,
>=20
> Friday, March 14, 2008, 6:48:52 PM, you wrote:
>=20
> > Hi!
>=20
> >> Lemme just think, doing inheritance at compile time before we send=20
> >> the stuff to an opcode cache can be slower then? How is that=20
> >> possible? After all late binding means we do it at run=20
> time. And no=20
> >> matter how much faster we can do it. It will always be slower than=20
> >> doing the same thing only once.
>=20
> > I'm taking this off-list because it doesn't seem to go any=20
> direction=20
> > it
> > can be useful for people to read.
>=20
> Hmm, so other people must not read what we discuss, that does=20
> not work.
>=20
> > Repeating again: it can not be done "only once". Not with opcode=20
> > caches
> > present and bound entities living in separate file. This=20
> thing is not=20
> > possible. If you do not believe it - try to implement it and see.
>=20
> That is my point. It can be done.=20

Look into the example:

parent1.php
<?php
class Foo {...}
?>

parent2.php
<?php
class Foo {...}
?>

child.php
<?php
class Bar extens Foo {...}
?>

index1.php
<?php
include "parent1.php";
include "child.php";
?>

index2.php
<?php
include "parent2.php";
include "child.php";
?>

Do you still think if early-binding for class Bar is possible at
compile-time with opcode caches?
And what "Foo" is the parent?

> The reason to delay=20
> inheritance is that we can have a conditional statement=20
> around a class definition.=20

There is no "if' statement in example above.
Also detection of conditional code is additional pine for caches.

> If we detect classes that do not=20
> have conditionals around them, then we can bind them early.

It's wrong.
=20
> It is as easy as that. So instead of making the language more=20
> dynamic and even removing a chance of having inheritance=20
> dones early where possible we should work on making the=20
> desired case - eraly binding - as easy as possible.

The proposed patch allows caches to implement proper behavior of
early-biniding in most easy way. (~5 lines of code)

Thanks. Dmitry.