From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 26818 invoked by alias); 25 Aug 2005 16:30:39 -0000 Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-bugs-owner@gcc.gnu.org Received: (qmail 26498 invoked by uid 48); 25 Aug 2005 16:29:55 -0000 Date: Thu, 25 Aug 2005 16:36:00 -0000 Message-ID: <20050825162955.26497.qmail@sourceware.org> From: "rth at gcc dot gnu dot org" To: gcc-bugs@gcc.gnu.org In-Reply-To: <20050825153435.23561.jakub@gcc.gnu.org> References: <20050825153435.23561.jakub@gcc.gnu.org> Reply-To: gcc-bugzilla@gcc.gnu.org Subject: [Bug rtl-optimization/23561] nonoverlapping_memrefs_p returns true even for overlapping memory references X-Bugzilla-Reason: CC X-SW-Source: 2005-08/txt/msg02928.txt.bz2 List-Id: ------- Additional Comments From rth at gcc dot gnu dot org 2005-08-25 16:29 ------- Careful, Andrew. Things are not as cut-and-dried as you're making it out. Indeed, this is yet another example of the big structure member aliasing discussion we had earlier this year. I can't find the reference just now. I vaguely seem to recall that for the nonce we decided to allow up-casts from members back to the structure. Which suggests that our string builtins should adjust the MEM_EXPR to discard the member from the aliasing info, and refer only to the object. -- What |Removed |Added ---------------------------------------------------------------------------- CC| |mark at codesourcery dot com http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23561