From mint-bounce@lists.fishpool.fi Sat Feb 14 18:36:59 2009 Message-ID: <49975513.9060606@atari-source.org> Date: Sat, 14 Feb 2009 18:34:43 -0500 From: Mark Duckworth User-Agent: Thunderbird 2.0.0.19 (Macintosh/20081209) MIME-Version: 1.0 To: =?ISO-8859-1?Q?Vincent_Rivi=E8re?= CC: mint Subject: Re: [MiNT] C++ Stuff References: <499665F5.809@atari-source.org> <4996E70E.1030603@atari-source.org> <4996EE34.8090002@freesbee.fr> <499740F4.2010203@atari-source.org> <49974575.2000506@freesbee.fr> In-Reply-To: <49974575.2000506@freesbee.fr> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit X-SA-Score: -1.4 X-ecartis-version: Ecartis v1.0.0 Sender: mint-bounce@lists.fishpool.fi Errors-to: mint-bounce@lists.fishpool.fi X-original-sender: mduckworth@atari-source.org Precedence: bulk List-help: List-unsubscribe: List-Id: X-List-ID: List-subscribe: List-owner: List-post: Vincent Rivière wrote: > Mark Duckworth wrote: >> It did crash (sort of): >> >> root@arawork:/h>g++ -o vect vect.cpp >> root@arawork:/h>./vect >> Not crashed ! screenname = >> >> It freezes here and if I try to attach to it with gdb I cannot >> interrupt execution (or maybe I'm doing it wrong). > > Humm... > You didn't answer the questions on my first message. > > 1) Which version of g++ do you use ? > Try g++ --version and tell me what it says. > I use the cross g++ 4.3.3 (and the brand-new binutils 2.19.1) and it > works fine. > > 2) Try to increase the size of the stack of the generated executable, > using stack -S > While you use a native MiNT compiler, it may be a problem with the > stack size of the compiler itself. It is something like > /usr/libexec/gcc/m68k-atari-mint/*/cc1plus > > You may have a look at the preprocessed output (using g++ -E vect.cpp) > and the assembler output (using g++ -S -o - vect.cpp), too, but it may > be quite long and complicated. > > Good luck. > Ok some more testing. First my version is 2.95.3 which is included with sparemint. I started out with this version because I figured it worked fine. Apparently it doesn't. The cc1plus executable has a 960KB stack which seems plenty high enough. I forgot I replaced the libstdc++ lib with the m68020-60 version. I put it back to the 68000 version and I don't get even as far as I did before. It simply just kills aranym with a bus error no matter what I set the stack to (32K, 64, etc). I wouldn't know how to look at the assembler output and make any sense of it. I could post it if someone else cares - but does it really matter? Maybe I should just upgrade to the later versions and we should stop using 2.95.3 already. Thanks, Mark