From mboxrd@z Thu Jan 1 00:00:00 1970 From: Igor Seleznev To: egcs-bugs@cygnus.com Subject: Bug report Date: Wed, 31 Mar 1999 23:54:00 -0000 Message-id: X-SW-Source: 1999-03n/msg00077.html List-Id: Hi, I think I have found a compiler bug, look in attachment for details. There are two C++ source files - 'test.cc' and 'test_new.cc' And there are only one difference between them: 10c10 < operator int() const { printf( "int operator\n" ); return 1; } --- > operator int() { printf( "int operator\n" ); return 1; } But it is enough for wrong work. For the first case it writes: bool operator int operator for the second file it writes: int operator int operator Seems that when I am calling an operator with 'const' modifier and there are one more operator exist without any 'const', then it will call last one instead of what I am wishing to be called. Regards, Igor Seleznev >>From loewis@informatik.hu-berlin.de Wed Mar 31 23:54:00 1999 From: Martin von Loewis To: forge@diku.dk Cc: egcs-bugs@egcs.cygnus.com Subject: Re: Bug report (fwd) Date: Wed, 31 Mar 1999 23:54:00 -0000 Message-id: <199903251625.RAA11549@pandora> References: X-SW-Source: 1999-03n/msg00829.html Content-length: 1283 > Upon running the testsuite (C and C++) I get queasy, however. A lot of the > tests fail. The installation instructions on the egcs homepage does say: [...] > ... but should I just ignore the 'FAIL: '-messages. An expert for the tests and for your platform is needed to answer that question. For example, it is often reported that, say, all exeception-handling patches fail. It can't hurt to pick a specific test that failed, and analyze why it failed; testsuite/*.log has all the details. > Also, I am having trouble debugging egcs-generated code. I have tried all > the different switches (-g, -ggdb, -gstabs, -gstabs+), but gdb (4.16) > cannot even set a breakpoint at the line, which calls sort, in the simple > program, which I have also attached. I cannot seem to find a remedy for > this in FAQs, mail-archives etc. A could try installing 4.17 of gdb, but I > feel that the problems go deeper than that. With a bit of bad luck, you've picked up a snapshot where debugging is broken. Please try to compile a file with '-S' and '-S -g'. If the assembler output shows no difference, you need a snapshot that has the change Sun Mar 21 17:33:48 1999 Jeffrey A Law (law@cygnus.com) * toplev.c (main): A debug option without a level defaults to level 2. Regards, Martin