From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 31023 invoked by alias); 11 Oct 2004 18:02:18 -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 31010 invoked by alias); 11 Oct 2004 18:02:18 -0000 Date: Mon, 11 Oct 2004 18:02:00 -0000 Message-ID: <20041011180218.31009.qmail@sourceware.org> From: "dave at hiauly1 dot hia dot nrc dot ca" To: gcc-bugs@gcc.gnu.org In-Reply-To: <20041011151947.17933.danglin@gcc.gnu.org> References: <20041011151947.17933.danglin@gcc.gnu.org> Reply-To: gcc-bugzilla@gcc.gnu.org Subject: [Bug rtl-optimization/17933] [4.0 Regression] ICE: in schedule_insns, at /sched-rgn.c:2555 X-Bugzilla-Reason: CC X-SW-Source: 2004-10/txt/msg01465.txt.bz2 List-Id: ------- Additional Comments From dave at hiauly1 dot hia dot nrc dot ca 2004-10-11 18:02 ------- Subject: Re: [4.0 Regression] ICE: in schedule_in > Here is a C example which is almost equivalent to the objective-C example: > struct d > { > int a; > }; > void abort(void); > typedef struct d (*f) (int i); > f ff(void); > void test1(void) > { > if ((ff())(0).a != 0) > abort(); > } It generates the same error. It seems like the problem is in count_or_remove_death_notes. There are 30 REG_DEAD notes in the objective-C example. Dave -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17933