Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:82105
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)
Message-ID: <54D6AF67.7050003@gmx.de>
Date: Sun, 08 Feb 2015 01:35:51 +0100
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:31.0) Gecko/20100101 Thunderbird/31.4.0
MIME-Version: 1.0
To: Oleg Serov <serovov@gmail.com>, internals@lists.php.net
References: <CAMkKa8BCxprjDhe=gNo=xdvw_bxtZFM_rPu7HZJEsqAaf64K1A@mail.gmail.com>
In-Reply-To: <CAMkKa8BCxprjDhe=gNo=xdvw_bxtZFM_rPu7HZJEsqAaf64K1A@mail.gmail.com>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Subject: Re: Idea of optimizing php !empty(...) expression.
From: cmbecker69@gmx.de (Christoph Becker)

Oleg Serov wrote:

> I use !empty() very often and decided to make a benchmark test.
> 
> Here is the code and results: http://pastebin.com/fMhhdQiW
> 
> if (!empty(...)) working on 23% slower than if (empty()) expression.
> 
> So if create new operator not_empty() it will improve performance.
> 
> The first question is: What do you think about optimizing !empty(...), do
> we need it ?
> 
> And I see two way to make this happen.
> 
> 1. Create new language entity "not_empty".
> 2. Improve parser and help to handle "!empty" calls different way.
> 
> It is obviously that option 2 is better. Is it real to optimize parser that
> way?

I see a third way: optimize this on the OPcode level, i.e. BOOL_NOT(X) +
JMPZ(T) -> NOP, JMPNZ(X).  That is already done by OPcache[1].

[1] <http://lxr.php.net/xref/PHP_5_5/ext/opcache/Optimizer/block_pass.c#816>

-- 
Christoph M. Becker