Newsgroups: php.internals Path: news.php.net Xref: news.php.net php.internals:112282 Return-Path: Delivered-To: mailing list internals@lists.php.net Received: (qmail 82416 invoked from network); 24 Nov 2020 21:39:33 -0000 Received: from unknown (HELO php-smtp4.php.net) (45.112.84.5) by pb1.pair.com with SMTP; 24 Nov 2020 21:39:33 -0000 Received: from php-smtp4.php.net (localhost [127.0.0.1]) by php-smtp4.php.net (Postfix) with ESMTP id C8FAD1804B3 for ; Tue, 24 Nov 2020 13:05:18 -0800 (PST) X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on php-smtp4.php.net X-Spam-Level: X-Spam-Status: No, score=-0.7 required=5.0 tests=BAYES_20,DKIM_SIGNED, DKIM_VALID,RCVD_IN_DNSWL_LOW,RCVD_IN_MSPIKE_H3,RCVD_IN_MSPIKE_WL, SPF_HELO_PASS,SPF_NONE autolearn=no autolearn_force=no version=3.4.2 X-Spam-Virus: No X-Envelope-From: Received: from wout2-smtp.messagingengine.com (wout2-smtp.messagingengine.com [64.147.123.25]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by php-smtp4.php.net (Postfix) with ESMTPS for ; Tue, 24 Nov 2020 13:05:17 -0800 (PST) Received: from compute4.internal (compute4.nyi.internal [10.202.2.44]) by mailout.west.internal (Postfix) with ESMTP id BBFF5A0D for ; Tue, 24 Nov 2020 16:05:16 -0500 (EST) Received: from imap26 ([10.202.2.76]) by compute4.internal (MEProxy); Tue, 24 Nov 2020 16:05:16 -0500 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d= messagingengine.com; h=content-type:date:from:in-reply-to :message-id:mime-version:references:subject:to:x-me-proxy :x-me-proxy:x-me-sender:x-me-sender:x-sasl-enc; s=fm1; bh=mMC+BH x4L68XMnF/s023QIAx6ZRjR0fETlAz5uF13rI=; b=goOs6zoc6tzDGfnbauUS3q 4OVhgU7chRXQ1kJ3rzGKgL3x+VyUGhog4zs/8cbE29tM9QY3XV4E3fNK9SXfmNXC RWFpTsCD7HJcQb1d0Jmy8I9M5djM1W7lJsqgbptl/VjixqQamL2WiYXEwUC+2loe bxUMAXi7wSRJ5PkW8rCT5UQS3747h583zghnyVF7nAoEyrsu3RVX10hWz/tm0eyD njo0E/H9Hwgk+XgNkn5CBAj7pe7DOl0LOhYdl7mGQMu4Itt39LfujMCUyRAHlVYP 5Dk66hnPhtkeRMXa7jwE1pqWl76facpEcBEWZDTQsNYu5VlWH48qdn+GEwZ4b47g == X-ME-Sender: X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgedujedrudegkedgudegiecutefuodetggdotefrod ftvfcurfhrohhfihhlvgemucfhrghsthforghilhdpqfgfvfdpuffrtefokffrpgfnqfgh necuuegrihhlohhuthemuceftddtnecusecvtfgvtghiphhivghnthhsucdlqddutddtmd enucfjughrpefofgggkfgjfhffhffvufgtsehttdertderredtnecuhfhrohhmpedfnfgr rhhrhicuifgrrhhfihgvlhgufdcuoehlrghrrhihsehgrghrfhhivghlughtvggthhdrtg homheqnecuggftrfgrthhtvghrnhepfeetveffteetueeukeegjeffudeuhedthfevfeeh iefhheegheffhedthefgleejnecuffhomhgrihhnpehphhhprdhnvghtpdhgihhthhhusg drtghomhenucevlhhushhtvghrufhiiigvpedtnecurfgrrhgrmhepmhgrihhlfhhrohhm pehlrghrrhihsehgrghrfhhivghlughtvggthhdrtghomh X-ME-Proxy: Received: by mailuser.nyi.internal (Postfix, from userid 501) id 0D59A14200A2; Tue, 24 Nov 2020 16:05:16 -0500 (EST) X-Mailer: MessagingEngine.com Webmail Interface User-Agent: Cyrus-JMAP/3.3.0-622-g4a97c0b-fm-20201115.001-g4a97c0b3 Mime-Version: 1.0 Message-ID: <9e902c98-8d97-4401-9dd1-29607dd6eff8@www.fastmail.com> In-Reply-To: References: Date: Tue, 24 Nov 2020 15:04:18 -0600 To: "php internals" Content-Type: text/plain Subject: Re: [PHP-DEV] PHP releases, OPcache + Jit bugs, and communication From: larry@garfieldtech.com ("Larry Garfield") On Tue, Nov 24, 2020, at 2:47 PM, Dan Ackroyd wrote: > Hi internals, > > Currently the PHP project doesn't have a particularly great way of > letting users know when serious defects have been found in versions of > PHP. > > My understanding is that this has been an issue before, when defects > were found in OPcache. Due to OPcache incorrectly optimizing code, > bugs could spontaneously appear anywhere in users code. As we had > nothing in place, we didn't have a way of communicating 'the latest > version is borked, avoid it' Fortunately there were few incidents of > this. > > However, the JIT is quite likely to have many similar issues, where > either new issues, or regressions, could seriously affect the > integrity of how data is processed in PHP applications. > > I'd like to suggest that this could be improved by having some machine > readable data somewhere (see example below), that contains a list of > known critical issues that people should know about before upgrading > to a particular version of PHP. > > This would at least allow people to either hold off on upgrading from > a version that works, to a known bad version, as well as do things > like alert their ops team of investigating whether a newly found issue > could be affecting their programs, and it might be appropriate for > them to revert to a previous version of PHP. > > Thoughts? And does anyone know of any projects that already do this, > so we can be inspired by their best practices? > > cheers > Dan > Ack > > btw before anyone suggests "why don't we just have more releases?", > PHP is mostly distributed through package managers on a fixed > schedule. Switching to an ad-hoc schedule would be a huge amount of > work for many people, and doesn't like a reasonable thing to do. > > > Example of data > --------------------- > [ > { > "version": "8.0.1", > "issues": [ > { > "link": "https:\/\/bugs.php.net\/bug.php?id=12345", > "affects": "jit" > } > ] > }, > { > "version": "8.0.0", > "issues": [ > { > "link": "https:\/\/bugs.php.net\/bug.php?id=12345", > "affects": "opcache" > } > ] > } > ] > > The 'affects' entry could be a comma separated list of things such as: > > jit - the JIT > opcache - opcache > php - the core engine with/without JIT or OPcache. > security - known security flaws that of a severity that justify an > urgent upgrade That was essentially the idea behind the FIG's PSR-9: https://github.com/php-fig/fig-standards/blob/master/proposed/security-disclosure-publication.md It unfortunately never really went anywhere, but I thought it was a good idea. There's some links there to some prior art we were drawing from, or planning to draw from. The idea was to allow projects to publish a link to a feed of security releases in their composer.json, and then Composer (or a plugin) could audit your dependencies and tell you if one of them was busted. My ideal was an Atom feed, as then it's compatible with pub/sub, but not everyone agreed with me about that. :-) --Larry Garfield