Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:95813
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmx.de designates 212.227.17.21 as permitted sender)
To: Andrea Faulds <ajf@ajf.me>, internals@lists.php.net
References: <F4.52.16256.34F1A275@pb1.pair.com>
 <9203225f-3367-6f31-7aa7-2a8ef373c858@gmx.de>
 <5E.70.06456.AB130D75@pb1.pair.com>
Message-ID: <8fbf9eb2-82b8-7b9a-1207-891a8c4546e5@gmx.de>
Date: Fri, 9 Sep 2016 00:20:59 +0200
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:45.0) Gecko/20100101
 Thunderbird/45.2.0
MIME-Version: 1.0
In-Reply-To: <5E.70.06456.AB130D75@pb1.pair.com>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Subject: Re: The death of `#ifndef FAST_ZPP`?
From: cmbecker69@gmx.de ("Christoph M. Becker")

Hi Andrea!

On 07.09.2016 at 17:26, Andrea Faulds wrote:

> Christoph M. Becker wrote:
>
>> On 04.05.2016 at 18:11, Andrea Faulds wrote:
> 
> This is a fair point. Functions using Fast ZPP have larger code size,
> because of aggressive inlining.
> 
> If every function used it, I suppose it might bloat the size of the PHP
> binary somewhat.

And probably for no good reason.  The performance overhead of
zend_parse_parameters() can probably be ignored for a lot of functions,
and while it is somewhat harder to read, besides it's being
space-saving, it is way easier to debug.

There are, however, functions that would greatly benefit from using the
fast ZPP API (e.g. imagecolorat(), see commit 0bf11d1e), so I'd welcome
code reviews regarding (fast) ZPP.

> In any case, thanks to some uncaught bugs in dead code, no stable
> release of PHP 7 has been able to compile with FAST_ZPP disabled anyway,
> which means I can be certain nobody is using it.

I have successfully compiled a recent master with FAST_ZPP undefined.
While that was with `--disable-all`, I can imagine that some extensions
would also compile, so perhaps somebody actually uses that feature.
However, even if that is the case, I don't see necessity to do so.

-- 
Christoph M. Becker