Message-ID: <250B3114DA16D511B82C00E0094005F8023FC195@MSGWAW11> From: Wojciech Galazka To: "'Eli Zaretskii'" Cc: djgpp-workers AT delorie DOT com Subject: RE: gcc-3.01 seems unstable Date: Fri, 21 Sep 2001 18:49:10 +0200 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Reply-To: djgpp-workers AT delorie DOT com > -----Original Message----- > From: Eli Zaretskii [mailto:eliz AT is DOT elta DOT co DOT il] > > > From: =?windows-1250?Q?Wojciech_Ga=B3=B9zka?= > > > Date: Thu, 20 Sep 2001 18:52:16 +0200 > > > > I have two separate directories \djgpp and \djgpp.2 and I > manually set the > > PATH and DJGPP variables > > I guessed that much. You may wish to see whether something in this > setup causes some mess. Are you sure all the rest of these two > directories, besides the compiler and its related files, is identical? > The \djgpp.2 directory contains files from clio (make, bash, and so on) Well, Today I tried to reproduce the crashes I saw before and the gcc-3.01 compiled the files in \djgpp\tests\cygnus with no problems. Since the tests are made on a laptop with some external devices connected to it, I tried different hardware configurations (including the same as when the crash was occuring) and it still compiled correctly Then I tried to compile the gcc-3.01 itself assuming that if it were to crash it will. I got several 'Segmentation violation, internal compiler error ...' messages but was going on ontil I got (this is in stage 2) ../../gcc/libgcc2.c: In function `__muldi3': ../../gcc/libgcc2.c:368: Internal compiler error in remove_unnecessary_notes, at emit-rtl.c:2897 Please submit a full bug report, with preprocessed source if appropriate. and stopped there I'm not sure, perhaps this is bad memory problem but could you check the following file (I got the gcc-3.01 sources from Andris' site) gnu\gcc-3.01\gcc\cp\rtti.c, line 848 which looks like (the asterisk should be read as a char of ASCII code 127) SET_DECL_ASSEMBLER_NAME (name*decl, when it should look like SET_DECL_ASSEMBLER_NAME (name_decl, I'm trying to check the compilation under W2K ... I