From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 19379 invoked by alias); 11 Oct 2005 13:51:34 -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 19348 invoked by uid 48); 11 Oct 2005 13:51:31 -0000 Date: Tue, 11 Oct 2005 13:51:00 -0000 Message-ID: <20051011135131.19347.qmail@sourceware.org> X-Bugzilla-Reason: CC References: Subject: [Bug rtl-optimization/23585] [4.0 regression] mem_fun* code fine with -O1, bus error with -O2 In-Reply-To: Reply-To: gcc-bugzilla@gcc.gnu.org To: gcc-bugs@gcc.gnu.org From: "ebotcazou at gcc dot gnu dot org" X-SW-Source: 2005-10/txt/msg01299.txt.bz2 List-Id: ------- Comment #8 from ebotcazou at gcc dot gnu dot org 2005-10-11 13:51 ------- > Yes, I think the problem is in delay slot scheduling too. COND_EXPR means that > either branch must not be evaluated because it could be illegal; if you hoist a > mem from a branch into the delay slot of the condition, you are effectively > partially evaluting the branch. Yes, that makes sense. At the moment the predicate used in reorg.c is may_trap_p so we would need to detect faulting instructions too. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23585