From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 17180 invoked by alias); 27 May 2007 14:50:35 -0000 Received: (qmail 17171 invoked by uid 22791); 27 May 2007 14:50:34 -0000 X-Spam-Check-By: sourceware.org Received: from stelecom.gomel.by (HELO stelecom.gomel.by) (82.209.213.61) by sourceware.org (qpsmtpd/0.31) with ESMTP; Sun, 27 May 2007 14:50:32 +0000 Received: from localhost (localhost [127.0.0.1]) by stelecom.gomel.by (Postfix) with ESMTP id 2775CB008C1F for ; Sun, 27 May 2007 17:50:30 +0300 (EEST) Received: from stelecom.gomel.by ([127.0.0.1]) by localhost (stelecom.gomel.by [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 30954-34 for ; Sun, 27 May 2007 17:50:29 +0300 (EEST) Received: from localhost (unknown [86.57.140.7]) by stelecom.gomel.by (Postfix) with ESMTP id 7FC56B010475 for ; Sun, 27 May 2007 17:50:28 +0300 (EEST) Date: Sun, 27 May 2007 23:51:00 -0000 From: Sergei Gavrikov To: ecos-discuss@ecos.sourceware.org Message-ID: <20070527144943.GA23557@ubuntu> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.5.13 (2006-08-11) X-IsSubscribed: yes Mailing-List: contact ecos-discuss-help@ecos.sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: ecos-discuss-owner@ecos.sourceware.org Subject: [ECOS] synth.ld (GCC 3.2.1) question X-SW-Source: 2007-05/txt/msg00138.txt.bz2 Hello, Sometimes, I use the i386-elf-gcc (GCC 3.2.1) from the public ecoscentric toolchain for I386 to build/debug some test using the eCos synth target. AFAIK, they don't recommend to use native GCC for that. The synth.ld script (hal/synth/arch/current/src/synth.ld) contains a point #if (__GNUC__ >= 3) GROUP(libtarget.a libgcc.a libsupc++.a libgcc_eh.a) #else GROUP(libtarget.a libgcc.a) #endif According a ChangeLog, that was Jonathan Larmour's tweak. But their GCC 3.2.1 has no libgcc_eh.a at all: ~/ecos/gnutools/i386-elf/i386-elf/lib/libcygmon.a ~/ecos/gnutools/i386-elf/i386-elf/lib/libsupc++.a ~/ecos/gnutools/i386-elf/i386-elf/lib/libstdc++.a ~/ecos/gnutools/i386-elf/i386-elf/lib/libnosys.a ~/ecos/gnutools/i386-elf/lib/gcc-lib/i386-elf/3.2.1/libgcc.a So, I have to fix the GROUP description in the result .ld script (install/lib/target.ld) every time to build the test, i.e. to replace the line GROUP(libtarget.a libgcc.a libsupc++.a libgcc_eh.a) by GROUP(libtarget.a libgcc.a libsupc++.a libgcc.a) Well, that isn't an issue for the sed, but... Question: Do we need to change that condition in the synth.ld to check the GCC minor version to fix the problem? Does somebody mess that too? The GCC spec is Reading specs from /home/sg/ecos/gnutools/i386-elf/bin/../lib/gcc-lib/i386-elf/3.2.1/specs Configured with: /home/demonweb/tools/ecos-gnutools-v1.4/r2/i386-elf/x86_linux_lsb1_3/tar_bz2/source/gcc-3.2.1/configure --target=i386-elf --prefix=/home/demonweb/tools/ecos-gnutools-v1.4/r2/i386-elf/x86_linux_lsb1_3/tar_bz2/opt/ecos/gnutools/i386-elf --enable-languages=c,c++ --with-gnu-as --with-gnu-ld --with-newlib --with-gxx-include-dir=/home/demonweb/tools/ecos-gnutools-v1.4/r2/i386-elf/x86_linux_lsb1_3/tar_bz2/opt/ecos/gnutools/i386-elf/i386-elf/include Thread model: single gcc version 3.2.1 The ecosconfig version is ecosconfig 2.0 (May 9 2003 09:45:47) Copyright (c) 2002 Red Hat, Inc. My usual steps to build any test are 1. Build libtarget.a ecosconfig new linux