From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 21668 invoked by alias); 21 Aug 2009 18:51:17 -0000 Received: (qmail 21659 invoked by uid 22791); 21 Aug 2009 18:51:17 -0000 X-SWARE-Spam-Status: No, hits=-2.4 required=5.0 tests=AWL,BAYES_00,J_CHICKENPOX_13,J_CHICKENPOX_55,RCVD_IN_DNSWL_LOW,SPF_PASS X-Spam-Check-By: sourceware.org Received: from eir.is.scarlet.be (HELO eir.is.scarlet.be) (193.74.71.27) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Fri, 21 Aug 2009 18:51:10 +0000 Received: from [172.17.1.10] (ip-81-11-241-121.dsl.scarlet.be [81.11.241.121]) by eir.is.scarlet.be (8.14.2/8.14.2) with ESMTP id n7LIokFU006207; Fri, 21 Aug 2009 20:50:46 +0200 Subject: Re: Build question From: Danny Backx Reply-To: danny.backx@scarlet.be To: Eli Zaretskii Cc: tromey@redhat.com, gdb@sourceware.org In-Reply-To: <83ab1tgh9h.fsf@gnu.org> References: <1250803105.11282.96.camel@pavilion> <83d46pgjkq.fsf@gnu.org> <1250877901.11282.116.camel@pavilion> <83ab1tgh9h.fsf@gnu.org> Content-Type: text/plain Date: Fri, 21 Aug 2009 20:15:00 -0000 Message-Id: <1250880746.11282.128.camel@pavilion> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit X-DCC-scarlet.be-Metrics: eir 20001; Body=4 Fuz1=4 Fuz2=4 Mailing-List: contact gdb-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-owner@sourceware.org X-SW-Source: 2009-08/txt/msg00216.txt.bz2 On Fri, 2009-08-21 at 21:21 +0300, Eli Zaretskii wrote: > > > The way these macros and the corresponding source fragments in GDB are > > > set, they only DTRT for native debugging. > > > > Also right, but it looks like it doesn't take much to make it just work. > > I think they need to be converted to run-time tests instead. For > that, we would probably need some kind of primitive that returns the > filesystem type of the target, what the macros do now at compile-time. That doesn't look right. I am building i386-mingw32ce-gdb (or arm-mingw32ce-gdb). The targets specified are Windows CE based, so they all have this "DOS BASED" file system. So why test this at run time ? The total extent of the patches I needed to get this particular issue out of the way is this : pavilion: {868} cvs diff include/filenames.h Index: include/filenames.h =================================================================== RCS file: /cvs/src/src/include/filenames.h,v retrieving revision 1.5 diff -u -r1.5 filenames.h --- include/filenames.h 21 Mar 2008 23:40:18 -0000 1.5 +++ include/filenames.h 21 Aug 2009 18:41:44 -0000 @@ -36,6 +36,10 @@ #define HAVE_DOS_BASED_FILE_SYSTEM 1 #endif +#endif + +#ifdef HAVE_DOS_BASED_FILE_SYSTEM + #define IS_DIR_SEPARATOR(c) ((c) == '/' || (c) == '\\') /* Note that IS_ABSOLUTE_PATH accepts d:foo as well, although it is only semi-absolute. This is because the users of IS_ABSOLUTE_PATH And in the build directory, I added #define HAVE_DOS_BASED_FILE_SYSTEM 1 to libiberty/config.h and gdb/config.h . I am looking for a way to get this into the build system somehow. Danny -- Danny Backx ; danny.backx - at - scarlet.be ; http://danny.backx.info