From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 16713 invoked by alias); 24 May 2010 20:40:23 -0000 Received: (qmail 16696 invoked by uid 22791); 24 May 2010 20:40:23 -0000 X-SWARE-Spam-Status: No, hits=-1.3 required=5.0 tests=AWL,BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FROM X-Spam-Check-By: sourceware.org Received: from mail-fx0-f41.google.com (HELO mail-fx0-f41.google.com) (209.85.161.41) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Mon, 24 May 2010 20:40:18 +0000 Received: by fxm11 with SMTP id 11so2096698fxm.0 for ; Mon, 24 May 2010 13:40:15 -0700 (PDT) Received: by 10.223.27.151 with SMTP id i23mr5070783fac.104.1274733614405; Mon, 24 May 2010 13:40:14 -0700 (PDT) Received: from [192.168.0.10] (dra38-7-88-179-85-53.fbx.proxad.net [88.179.85.53]) by mx.google.com with ESMTPS id 15sm21229226fad.10.2010.05.24.13.40.12 (version=SSLv3 cipher=RC4-MD5); Mon, 24 May 2010 13:40:12 -0700 (PDT) Subject: Re: [commit] updated gnulib from latest git. From: =?ISO-8859-1?Q?Fr=E9d=E9ric?= RISS To: Pedro Alves Cc: gdb-patches@sourceware.org In-Reply-To: <201005230240.12389.pedro@codesourcery.com> References: <201005230240.12389.pedro@codesourcery.com> Content-Type: text/plain; charset="UTF-8" Date: Mon, 24 May 2010 21:31:00 -0000 Message-ID: <1274733610.1918.15.camel@localhost.localdomain> Mime-Version: 1.0 Content-Transfer-Encoding: 8bit X-IsSubscribed: yes Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org X-SW-Source: 2010-05/txt/msg00553.txt.bz2 Hi, Le dimanche 23 mai 2010 à 02:40 +0100, Pedro Alves a écrit : > Following up from: > > > > I've updated gnulib from latest git. I don't understand all the specifics, but it seems there is something broken in this commit. The commit contains the followig hunk in gdb/config.in: -/* Number of bits in a file offset, on hosts where this is settable. */ -#undef _FILE_OFFSET_BITS - This means that on a 32bits host the GDB part of the tree will get compiled without large file support, while eg. bfd will be built with that support. This in turn means that gdb functions that pass 'struct stat *' to bfd functions will experience memory corruption. I've seen crashes in sepdebug.exp due to that corruption. Cheers, Fred