From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 113586 invoked by alias); 18 Apr 2018 13:43:26 -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 113575 invoked by uid 89); 18 Apr 2018 13:43:25 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-2.0 required=5.0 tests=AWL,BAYES_00,SPF_PASS autolearn=ham version=3.3.2 spammy=D*5, H*r:a9d X-HELO: mailbackend.panix.com X-Gm-Message-State: ALQs6tB1MqT52zk+4krX6uGmm9/s1HpmHybGIKkFhWTubOSjEliFwUDC CIXVBgedDrhN18rvOFCKxVeocxj6zt4CvkYZ4wY= X-Google-Smtp-Source: AIpwx4+cTEROvz/NFu16wVK7c14A9srdx8f4pNjJu6T4ILbei1aMJLG2WiKRv4Y2p8DasLb0L9AeVBemAB/Pqjysay0= X-Received: by 2002:aca:4f91:: with SMTP id d139-v6mr1236286oib.267.1524059001409; Wed, 18 Apr 2018 06:43:21 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: References: <20180416013614.29962-1-zackw@panix.com> From: Zack Weinberg Date: Wed, 18 Apr 2018 13:43:00 -0000 X-Gmail-Original-Message-ID: Message-ID: Subject: Re: [PATCH] New configure option --disable-libcrypt. To: Florian Weimer Cc: GNU C Library , "Carlos O'Donell" , Nikos Mavrogiannopoulos Content-Type: text/plain; charset="UTF-8" X-SW-Source: 2018-04/txt/msg00307.txt.bz2 On Tue, Apr 17, 2018 at 6:21 AM, Florian Weimer wrote: > On 04/16/2018 05:56 PM, Zack Weinberg wrote: > >> Let me see if I understand what you have in mind: In >> --disable-libcrypt mode, we would preserve _XOPEN_CRYPT, with value 1, >> and the prototype for crypt in unistd.h. > > My main proposal is preserving the prototype for crypt, under _GNU_SOURCE if > necessary (without defining _XOPEN_CRYPT). > > I just don't see value in the friction caused by dropping the definition. Makes sense. > It could still be a conformance violation for setkey/encrypt because current > libxcrypt (at least the version in Fedora) does not provide the functions > anymore for linking: > > $ eu-readelf --symbols=.dynsym /lib64/libcrypt.so.1 | grep encrypt > 38: 000000000000b330 20 FUNC GLOBAL DEFAULT 12 > encrypt_r@GLIBC_2.2.5 > 52: 000000000000b360 16 FUNC GLOBAL DEFAULT 12 > encrypt@GLIBC_2.2.5 Yeah, I did that on purpose; the total insecurity of single DES outweighs standard compliance in this case, I think. And, as you say, there probably aren't very many users. I'll send a revised patch shortly. zw