From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 30170 invoked by alias); 22 Aug 2008 17:55:24 -0000 Received: (qmail 28314 invoked by uid 48); 22 Aug 2008 17:54:03 -0000 Date: Fri, 22 Aug 2008 17:55:00 -0000 Message-ID: <20080822175403.28313.qmail@sourceware.org> X-Bugzilla-Reason: CC References: Subject: [Bug middle-end/36902] Array bound warning with dead code after optimization In-Reply-To: Reply-To: gcc-bugzilla@gcc.gnu.org To: gcc-bugs@gcc.gnu.org From: "manu at gcc dot gnu dot org" Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-bugs-owner@gcc.gnu.org X-SW-Source: 2008-08/txt/msg01700.txt.bz2 ------- Comment #21 from manu at gcc dot gnu dot org 2008-08-22 17:54 ------- (In reply to comment #19) > (In reply to comment #18) > > I think that if the compiler knows that the code is never executed then, we > > shouldn't warn. > > It does but only later on in the optimization it knows that the code is dead. My point is that if I understand the vrp code correctly, we can detect that the code is dead before warning. In this case, the code is detected dead within VRP, not in a later pass. Maybe I am wrong. Let's see what Mueller and Guenther have to say. In any case, we should consider whether warning in conditional BBs is a good idea at all. I think that saying "array subscript *is* above array bounds" in a BB that is executed conditionally is not a good idea in general. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36902