From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 16657 invoked by alias); 25 May 2016 18:25:28 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: libc-alpha-owner@sourceware.org Received: (qmail 16616 invoked by uid 89); 25 May 2016 18:25:27 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-3.3 required=5.0 tests=BAYES_00,RP_MATCHES_RCVD,SPF_HELO_PASS autolearn=ham version=3.3.2 spammy=cutting, brain, picture X-HELO: mx1.redhat.com Subject: Re: [PATCH v3] libio: use PTR_MANGLE/PTR_DEMANGLE for FILE vtables To: Kees Cook , libc-alpha@sourceware.org, Adam Conrad , Joseph Myers , Yunlian Jiang References: <20160524165854.GA32022@www.outflux.net> <0ebd00c0-7bbb-dbc5-d3e7-7fa6ba83c376@redhat.com> <20160525175245.GD26300@vapier.lan> From: Florian Weimer Message-ID: <2f4d8ed0-fc9f-0a5f-83aa-b28aaacd00bd@redhat.com> Date: Wed, 25 May 2016 18:43:00 -0000 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Thunderbird/45.0 MIME-Version: 1.0 In-Reply-To: <20160525175245.GD26300@vapier.lan> Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit X-SW-Source: 2016-05/txt/msg00594.txt.bz2 On 05/25/2016 07:52 PM, Mike Frysinger wrote: > On 25 May 2016 10:29, Florian Weimer wrote: >> I think we need a follow-up patch which removes the symbols which are >> now incompatible from the ABI, so that programs which are broken by this >> configuration receive an explicit error message. > > you mean all of the _IO_* symbols ? There are exceptions which are part of the API. At least _IO_putc and _IO_getc are, and _IO_stdin_used is interposed from the application. (We can maybe get rid of the latter because it does not work anyway because nothing ensures that the program is compiled against a consistent set of libio definitions.) > my familiarity here is like 0, Sadly, I don't know if we can pick anyone's brain who is more familiar with these details. > but are there any reasons to keep exporting those for new ABIs ? For most symbols, no, there is no reason to keep them for new ABIs. My understanding of symbol versioning is still limited, and I have no clear picture of the minimal libio ABI we can export, otherwise I'd have sent some patches long ago. A first step would involve cutting down the contents of the installed header file to the absolute minimum needed by . But even that is quite a bit of work. Florian