This is the mail archive of the libstdc++@sourceware.cygnus.com mailing list for the libstdc++ project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: libstdc++-2.90.4 and egcs-19990314 on i586-pc-linux-gnu


Hello Martin, and thanks again!

>>>>> "ML" == Martin v Loewis <martin@mira.isdn.cs.tu-berlin.de> writes:

ML> (honor-std is documented, BTW).

This is a good point. Er... There is no copy of the docs in the full
snapshots? I couldn't find it. Nightmare: there is a copy of the full
texinfo system!? I'll look again.

ML> The goal (AFAICT) is to provide a means to distribute experimental
ML> features whose use would break binary compatibility. What these exact
ML> features are might change from snapshot to snapshot; I guess this is
ML> the reason why they are not documented. Currently, it is a combination
ML> of:
ML> - a new layout of base classes, saving space when you can,
ML> - vtable thunks, if they are not enabled by default,
ML> - the squangling algorithm, and
ML> - the activation of std:: as a true namespace.

ML> Soon, it might also cover a change in the vtable thunks which is not
ML> fully compatible with the current implementation.

Thanks, but this kind of confirms what I thought... I am nobody to
make any critic, but this is not easy to get in!

ML> I'm somewhat lost here as to what exactly you are doing. I thought
ML> this is libstdc++ v3, where ostream should be a template, not a class.

Ahum. Good point indeed.

ML> Also, the correct symbol should be 

ML> std::ostream::operator<<(std::ostream &(*)(std::ostream &))

ML> Please compile your programs with --save-temps, then analyse the
ML> object files using nm(1); do not use nm's demangling option. Then
ML> invoke nm on the library are linking with.

ML> Without further information, I'd say you are getting the wrong header
ML> files. Please use --trace-includes or inspect the resulting .ii file
ML> to verify if that is the case.

Bingo :-/ The correct include directive was:

  -I/home/marc/H-libstdc++/include/g++-v3
                                   ^^^^^^

ML> Hope this helps,

It does...

% c++ -I/home/marc/H-libstdc++/include/g++-v3 -L/home/marc/H-libstdc++/lib -fnew-abi -fno-honor-std -o 27hello 27hello.cc
In file included from /home/marc/H-libstdc++/include/g++-v3/bits/std_iterator.h:34,
                 from /home/marc/H-libstdc++/include/g++-v3/bits/std_ios.h:53,
                 from /home/marc/H-libstdc++/include/g++-v3/bits/std_ostream.h:38,
                 from /home/marc/H-libstdc++/include/g++-v3/bits/std_iostream.h:38,
                 from /home/marc/H-libstdc++/include/g++-v3/iostream:2,
                 from 27hello.cc:21:
/home/marc/H-libstdc++/include/g++-v3/bits/stl_iterator_base.h:42: Internal compiler error 981117.
/home/marc/H-libstdc++/include/g++-v3/bits/stl_iterator_base.h:42: Please submit a full bug report to `egcs-bugs@egcs.cygnus.com'.
/home/marc/H-libstdc++/include/g++-v3/bits/stl_iterator_base.h:42: See <URL:http://egcs.cygnus.com/faq.html#bugreport> for details.
% 

(it doesn't go far enough to discriminate on whether I use std:: or
not)

One step at a time :-)

Thanks!
Marc

-- 
Marc Girod                Hiomo 5/1          Voice:  +358-9-511 23746
Nokia Telecommunications  P.O. Box 320       Mobile: +358-40-569 7954
NWS/NMS/NMS for Data      00045 NOKIA Group  Fax:    +358-9-511 23580
                          Finland            marc.girod@ntc.nokia.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]