From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 27560 invoked by alias); 19 Jun 2005 23:15:33 -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 27537 invoked by uid 48); 19 Jun 2005 23:15:28 -0000 Date: Sun, 19 Jun 2005 23:15:00 -0000 Message-ID: <20050619231528.27536.qmail@sourceware.org> From: "pinskia at gcc dot gnu dot org" To: gcc-bugs@gcc.gnu.org In-Reply-To: <20050417202952.21076.laurent@guerby.net> References: <20050417202952.21076.laurent@guerby.net> Reply-To: gcc-bugzilla@gcc.gnu.org Subject: [Bug tree-optimization/21076] [4.1 Regression] ACATs ICE cxh1001 at tree-vrp.c:124 X-Bugzilla-Reason: CC X-SW-Source: 2005-06/txt/msg02590.txt.bz2 List-Id: ------- Additional Comments From pinskia at gcc dot gnu dot org 2005-06-19 23:15 ------- (In reply to comment #3) > I think I fixed the bug which removed the non useless casts. Diego wasn't there a workaround in VRP > for this bug? I should say the following patch fixed the removing of the non useless casts: * tree-ssa.c (tree_ssa_useless_type_conversion_1): Conversions between integer types whos ranges are different are not useless. Which was PR 21923, we had got another ICE in VRP because of that. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21076