Newsgroups: php.internals Path: news.php.net Xref: news.php.net php.internals:70785 Return-Path: Mailing-List: contact internals-help@lists.php.net; run by ezmlm Delivered-To: mailing list internals@lists.php.net Received: (qmail 48494 invoked from network); 20 Dec 2013 14:18:32 -0000 Received: from unknown (HELO lists.php.net) (127.0.0.1) by localhost with SMTP; 20 Dec 2013 14:18:32 -0000 Authentication-Results: pb1.pair.com smtp.mail=ab@php.net; spf=unknown; sender-id=unknown Authentication-Results: pb1.pair.com header.from=ab@php.net; sender-id=unknown Received-SPF: unknown (pb1.pair.com: domain php.net does not designate 85.214.73.107 as permitted sender) X-PHP-List-Original-Sender: ab@php.net X-Host-Fingerprint: 85.214.73.107 klapt.com Received: from [85.214.73.107] ([85.214.73.107:43634] helo=klapt.com) by pb1.pair.com (ecelerity 2.1.1.9-wez r(12769M)) with ESMTP id 86/AE-63357-4B154B25 for ; Fri, 20 Dec 2013 09:18:31 -0500 Received: by klapt.com (Postfix, from userid 33) id 80D3A23D6137; Fri, 20 Dec 2013 15:18:25 +0100 (CET) Received: from 178.7.116.129 (SquirrelMail authenticated user anatol@belski.net) by webmail.klapt.com with HTTP; Fri, 20 Dec 2013 15:18:25 +0100 Message-ID: <3cf29b10143104230749852cd10a735f.squirrel@webmail.klapt.com> In-Reply-To: References: <2d2b8833cdd18298c7ecf19bac3ef3ea.squirrel@webmail.klapt.com> Date: Fri, 20 Dec 2013 15:18:25 +0100 To: "Martin Keckeis" Cc: internals@lists.php.net Reply-To: "Anatol Belski" User-Agent: SquirrelMail/1.5.2 [SVN] MIME-Version: 1.0 Content-Type: text/plain;charset=UTF-8 Content-Transfer-Encoding: 8bit Subject: Re: [PHP-DEV] 64 bit platform improvements From: ab@php.net ("Anatol Belski") Hi Martin, On Fri, December 20, 2013 09:52, Martin Keckeis wrote: > I'm having two open questions: > - is this the continued work where Anthony started? Or completely > restarted? That's the branch initiated by Anthony where we've worked on size_t changes only first. int64 came later and is now integrated with all together. > - does this patch have anything related with the current > "experimental" x64 > builds on windows.php.net? No. The "experimental" x64 branch uses 'int' for string size and 'long' for integer, which physically can hold 64 bit numbers on Windows. The str_size_and_int64 branch is a gigantic patch fixing those two things over all the PHP sources. That's why it needs to be good tested and reviewed. I haven't mention Windows builds, you can find todays snapshot here http://windows.php.net/downloads/snaps/ostc/str_size_and_int64/ . The autobuilds are under http://windows.php.net/downloads/snaps/str_size_and_int64/ (yet some older, will run again soon ). > > 2013/12/20 Anatol Belski > > >> Hi, >> >> >> I'm writing in the first place to give an update on the patch status. >> Just >> to remind - it's a matter of implementing size_t for string sizes and >> getting rid of 'long' in favour of platform dependent 64 bit integer. >> The >> patch is now close to finish the first big porting step. ZE and other >> very core parts are revamped in multiple circles and do now work well in >> the patch. Also the most important and wide used core extensions and >> SAPIs are >> ported. The progress page is reachable under >> https://wiki.php.net/rfc/string-size_t/progress , the RFC draft is >> linked from there. The work is of course still in progress. >> >> Regarding to my observations, there are some half dead or unsupported >> SAPIs and extensions currently in the core. Just to name some for SAPIs >> - >> phttpd, aolserver. >> >> Another observation is - there are 64 bit issues regarding dependency >> libraries. Though that was an open secret right at the start, many of >> those regarding 64 vs 32 bit integers do already persist in the 64 bit >> Linux/Unix PHP builds. >> >> >> Some steps are necessary to reach the next stage >> >> >> - strong code reviews >> - more testing >> - helping hands on porting >> >> >> In the light of the above, I would sincerely invite the community to >> participate on reviews, testing and further fixes. From the side of core >> devs, it would rock to have fixes for SAPIs and extensions rarely >> supported, especially the sapi/fpm wasn't ported yet. As well the >> ported stuff possibly (or better to say surely) has something overseen. >> Any >> improvement ideas are appreciated of course. >> >> On the part of people without C knowledge - testing with some real PHP >> apps would be already a great help. The usage is anyway interesting for >> needs like large strings/uploads, LFS, etc. Particular on Windows 64 >> bit integers/array indexes and big memory limit are enjoyable. Even >> running just the PHPTs on your machine were great. To compile yourself, >> no additional configure options are needed anymore. Even testing just >> one extension would help to chase down possible side effects and bugs. >> Not to >> forget, that the 32 bit builds of that branch should be tested, too. >> >> Thanks >> >> >> Anatol >> >> >> Best Regards Anatol