From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 10211 invoked by alias); 15 Oct 2003 14:30:47 -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 10196 invoked by uid 48); 15 Oct 2003 14:30:47 -0000 Date: Wed, 15 Oct 2003 14:30:00 -0000 Message-ID: <20031015143047.10195.qmail@sources.redhat.com> From: "rearnsha at gcc dot gnu dot org" To: gcc-bugs@gcc.gnu.org In-Reply-To: <20031006200259.12527.debian-gcc@lists.debian.org> References: <20031006200259.12527.debian-gcc@lists.debian.org> Reply-To: gcc-bugzilla@gcc.gnu.org Subject: [Bug target/12527] [3.4 regression] [arm] bootstrap error on arm-linux, miscompiling genconstants X-Bugzilla-Reason: CC X-SW-Source: 2003-10/txt/msg01064.txt.bz2 List-Id: PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org. http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12527 rearnsha at gcc dot gnu dot org changed: What |Removed |Added ---------------------------------------------------------------------------- Status|WAITING |NEW Last reconfirmed|2003-10-14 04:10:30 |2003-10-15 14:30:46 date| | ------- Additional Comments From rearnsha at gcc dot gnu dot org 2003-10-15 14:30 ------- I've just realized what is going wrong here. It's because the arm-linux target is (erroneously in my opinion) including arm/unknown-elf.h. This file defaults the CPU target to ARM7TDMI, which means that the bx instruction becomes available. The patch in question just means that this latent bug has gone critical. You can probably work around this problem for now by configuring with --with-cpu=arm6 or some other suitable CPU.