From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on inbox.vuxu.org X-Spam-Level: X-Spam-Status: No, score=-3.0 required=5.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,RCVD_IN_DNSWL_MED,RCVD_IN_MSPIKE_H3, RCVD_IN_MSPIKE_WL autolearn=ham autolearn_force=no version=3.4.2 Received: from mother.openwall.net (mother.openwall.net [195.42.179.200]) by inbox.vuxu.org (OpenSMTPD) with SMTP id c9f0d0b6 for ; Sun, 19 Jan 2020 19:13:18 +0000 (UTC) Received: (qmail 10216 invoked by uid 550); 19 Jan 2020 19:13:16 -0000 Mailing-List: contact musl-help@lists.openwall.com; run by ezmlm Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-ID: Reply-To: musl@lists.openwall.com Received: (qmail 10194 invoked from network); 19 Jan 2020 19:13:16 -0000 Date: Sun, 19 Jan 2020 22:13:05 +0300 (MSK) From: Alexander Monakov To: musl@lists.openwall.com In-Reply-To: <20200119173804.GH30412@brightrain.aerifal.cx> Message-ID: References: <20200119110743.GD2020@voyager> <20200119113134.GJ23985@port70.net> <8299f261-7870-57a6-37cf-d4ce482ad81e@openwall.com> <20200119161851.GC30412@brightrain.aerifal.cx> <20200119171908.GG30412@brightrain.aerifal.cx> <20200119173804.GH30412@brightrain.aerifal.cx> User-Agent: Alpine 2.20.13 (LNX 116 2015-12-14) MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Subject: Re: [musl] Minor style patch to exit.c On Sun, 19 Jan 2020, Rich Felker wrote: > > In the context of LTO asms really need to mention referenced symbols in > > constraints. Plus, it's good to have "the pointer is related to _start" > > spelled out in the code. > > I think this is true for objects defined in C code, but not for ones > defined in external asm or by the linker, which are not subject to any > analysis by LTO. Both the contents of inline asm statements and the > contents of external asm are black boxes, no? I think if they're not > treated as such lots of reasonable things would break. Yes, I was speaking in a more general sense. I agree here there would be no problem technically; I'd still insist on mentioning _start in constraints as a matter of clarity though. Alexander