From mint-bounce@lists.fishpool.fi Wed Feb 25 17:28:59 2009 Message-ID: <49A5C597.9030003@freesbee.fr> Date: Wed, 25 Feb 2009 23:26:31 +0100 From: =?ISO-8859-1?Q?Vincent_Rivi=E8re?= User-Agent: Thunderbird 2.0.0.19 (Windows/20081209) MIME-Version: 1.0 To: Mark Duckworth CC: mint@fishpool.com Subject: Re: [MiNT] Gcc 3.3.6 v 4.0.1 References: <20090225103817.swnyw26m0wgk8osc@pop.freeola.net> <373EB9E0-576D-4D59-B0A9-829AD39EB08F@gmail.com> <49A535ED.6000008@freesbee.fr> <49A56B1A.9050304@atari-source.org> In-Reply-To: <49A56B1A.9050304@atari-source.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed X-ecartis-version: Ecartis v1.0.0 Sender: mint-bounce@lists.fishpool.fi Errors-to: mint-bounce@lists.fishpool.fi X-original-sender: vincent.riviere@freesbee.fr Precedence: bulk List-help: List-unsubscribe: List-Id: X-List-ID: List-subscribe: List-owner: List-post: Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by mail.sparemint.org id n1PMSwZ2005728 Mark Duckworth wrote: > I'm stuck on gmp > right now. We can get past it by disabling gmp's use of obstack_vprintf I tried to cross-build libgmp with my latest GCC 4.3.3 on Cygwin, it has built fine. I have tricked the Makefiles to cross-build the tests (without running them), then I ran them on ARAnyM. (I don't know why but the C++ tests don't seem to have been built). Most of the test have passed without intervention. I had to adjust the stack for three of them: stack -S 256K t-fdiv stack -S 256K t-tdiv stack -S 256K reuse After that, they ran fine. Then I went to the t-printf test. I guess I got the same output as you: gmp_obstack_vprintf wrong fmt |%Md| got || want |0| got_len -1 ob_len 0 want_len 1 The test calls gmp_obstack_vprintf(), then it turns to call the MiNTLib in obstack_vprintf() then vprintf(). vfprintf() returns -1, it means failure. Look at the MiNTLib sources at mintlib/stdio/vfprintf.c http://sparemint.org/cgi-bin/cvsweb/mintlib/stdio/vfprintf.c?rev=1.5 That source is quite horrible and not debugger friendly :-( It seems that it fails very soon due to the macro ARGCHECK() (maybe because a bad FILE status ??) So maybe gmp tries to use an unsupported feature of the MiNTLib, or there is a bug in the MiNTLib, or something totally different. But I'm too tired for continuing now. -- Vincent Rivière