Linking problems, mangling ?

Linking problems, mangling ?

[Roland TTK]

 Hi
I have (succesfully ?) compiled my sources under solaris 2.6 with g++ 2.95 but
I get undefined reference errors when linking. I try to link my object files
with a library which was compiled by the sun C++ compiler. It seems like this
compiler uses a specific mangling grammar, so the linker can't find the
definitions it needs.

I think the mangling grammar is pretty old (1994) so I can't beleive gcc
doesn't emulate it.

Help !

Thanks for your inputs.

-- 
 TTK
 http://www.multimania.com/ttk22

Re: Linking problems, mangling ?

[Eric Meijer]

Roland TTK (hamon@timide.stna.dgac.fr.) wrote:
: 
:  Hi
: I have (succesfully ?) compiled my sources under solaris 2.6 with g++ 2.95 but
: I get undefined reference errors when linking. I try to link my object files
: with a library which was compiled by the sun C++ compiler.

You are trying to do something which cannot work.

: It seems like this
: compiler uses a specific mangling grammar, so the linker can't find the
: definitions it needs.

: I think the mangling grammar is pretty old (1994) so I can't beleive gcc
: doesn't emulate it.

You'd better believe it.  Quote from the g++ faq:

------------------------------------------------------------------------

Why can't g++ code link with code from other C++ compilers?

"Why can't I link g++-compiled programs against libraries compiled by
some other C++ compiler?"

Some people think that, if only the FSF and Cygnus Support folks would
stop being stubborn and mangle names the same way that, say, cfront
does, then any g++-compiled program would link successfully against
any cfront-compiled library and vice versa. Name mangling is the least
of the problems. Compilers differ as to how objects are laid out, how
multiple inheritance is implemented, how virtual function calls are
handled, and so on, so if the name mangling were made the same, your
programs would link against libraries provided from other compilers
but then crash when run. For this reason, the ARM encourages compiler
writers to make their name mangling different from that of other
compilers for the same platform. Incompatible libraries are then
detected at link time, rather than at run time.

------------------------------------------------------------------------

This means that you need to compile all C++ source in an application
with one and the same compiler.  With C libraries this problem generally
doesn't exist.

HTH,
Eric

-- 
 E.L. Meijer (tgakem@pebbles.chem.tue.nl)
 Eindhoven Univ. of Technology
 Lab. for Catalysis and Inorg. Chem. (SKA)

Re: Linking problems, mangling ?

[Roland TTK]

Eric Meijer a dit:
 >You are trying to do something which cannot work.
Yuk !

 >This means that you need to compile all C++ source in an application with
 >one and the same compiler.  With C libraries this problem generally doesn't
 >exist.
Thanks a lot for your help. But ...
I have to use these libraries and I don't have the source code so I can't
recompile them with gcc. So the solution should be to compile my sources with
the sun C++ compiler. But this one doesn't have the STL. So I tried to compile
the libstdc++ in order to emulate the STL.
Alas the sun compiler was unable to do it because libstdc++ sources contain
too many gcc specific things.

So I'm f---ed  :(

Or maybe I could find another implementation of the STL ??

Thanks.
-- 
 TTK
 http://www.multimania.com/ttk22

Re: Linking problems, mangling ?

[Mike Albaugh]

Eric Meijer (tgakem@pebbles.chem.tue.nl) wrote:

: [...], and so on, so if the name mangling were made the same, your
: programs would link against libraries provided from other compilers
: but then crash when run. For this reason, the ARM encourages compiler
: writers to make their name mangling different from that of other
: compilers for the same platform. Incompatible libraries are then
: detected at link time, rather than at run time.

: ------------------------------------------------------------------------

: This means that you need to compile all C++ source in an application
: with one and the same compiler.  With C libraries this problem generally
: doesn't exist.

	Beg to differ. On _many_ platforms, the C libraries from
different vendors (and even from different switch settings to the
same compiler) can and do differ in such things as parameter types,
structure layouts, etc. Heck MSVC's "intrinsic" functions disagree
with their own .h files, sometimes :-) Anyway, the result of these
"bleeding ends" is not typically detected at link time, but either
at compile time (_if_ the programmer cranks warnings up high enough
_and_ the .h files are complete and correct) or at run-time, where
they manifest themselves as obscure behavior.

						Mike
| albaugh@agames.com, speaking only for myself

Re: Linking problems, mangling ?

[Dave Sieber]

Roland TTK wrote:

> Or maybe I could find another implementation of the STL ??

Take a look at http://www.stlport.org/

--
dave

Re: Linking problems, mangling ?

[David C.]

hamon@timide.stna.dgac.fr. (Roland TTK) writes:
> 
> Or maybe I could find another implementation of the STL ??

Have you checked to see if you can get an updated compiler from Sun?
Their "Visual Workshop C++" package
( http://www.sun.com/workshop/visual/spec.html ) claims to be ISO
compilant, which means it should have STL support.

I suggest downloading a 30-day trial version (from Sun's web site) to
see if it's compatible with the library you want to use.  If it is, then
you may want to consider purchasing one or more licenses.

Unless/until your library vendor releases a gcc version, I don't think
you many other choices.  (If you're not using too much of the STL, you
might be able to write some quick-and-dirty implementations of the
templates you need, but that may end up being more work than it's
worth.)

-- David

Re: Linking problems, mangling ?

[David C.]

hamon@timide.stna.dgac.fr. (Roland TTK) writes:
> 
> Or maybe I could find another implementation of the STL ??

Have you checked to see if you can get an updated compiler from Sun?
Their "Visual Workshop C++" package
( http://www.sun.com/workshop/visual/spec.html ) claims to be ISO
compilant, which means it should have STL support.

I suggest downloading a 30-day trial version (from Sun's web site) to
see if it's compatible with the library you want to use.  If it is, then
you may want to consider purchasing one or more licenses.

Unless/until your library vendor releases a gcc version, I don't think
you many other choices.  (If you're not using too much of the STL, you
might be able to write some quick-and-dirty implementations of the
templates you need, but that may end up being more work than it's
worth.)

-- David

Re: Linking problems, mangling ?

[Eric Meijer]

Roland TTK (hamon@timide.stna.dgac.fr.) wrote:
: 
:  Hi
: I have (succesfully ?) compiled my sources under solaris 2.6 with g++ 2.95 but
: I get undefined reference errors when linking. I try to link my object files
: with a library which was compiled by the sun C++ compiler.

You are trying to do something which cannot work.

: It seems like this
: compiler uses a specific mangling grammar, so the linker can't find the
: definitions it needs.

: I think the mangling grammar is pretty old (1994) so I can't beleive gcc
: doesn't emulate it.

You'd better believe it.  Quote from the g++ faq:

------------------------------------------------------------------------

Why can't g++ code link with code from other C++ compilers?

"Why can't I link g++-compiled programs against libraries compiled by
some other C++ compiler?"

Some people think that, if only the FSF and Cygnus Support folks would
stop being stubborn and mangle names the same way that, say, cfront
does, then any g++-compiled program would link successfully against
any cfront-compiled library and vice versa. Name mangling is the least
of the problems. Compilers differ as to how objects are laid out, how
multiple inheritance is implemented, how virtual function calls are
handled, and so on, so if the name mangling were made the same, your
programs would link against libraries provided from other compilers
but then crash when run. For this reason, the ARM encourages compiler
writers to make their name mangling different from that of other
compilers for the same platform. Incompatible libraries are then
detected at link time, rather than at run time.

------------------------------------------------------------------------

This means that you need to compile all C++ source in an application
with one and the same compiler.  With C libraries this problem generally
doesn't exist.

HTH,
Eric

-- 
 E.L. Meijer (tgakem@pebbles.chem.tue.nl)
 Eindhoven Univ. of Technology
 Lab. for Catalysis and Inorg. Chem. (SKA)

Re: Linking problems, mangling ?

[Dave Sieber]

Roland TTK wrote:

> Or maybe I could find another implementation of the STL ??

Take a look at http://www.stlport.org/

--
dave

Re: Linking problems, mangling ?

[Mike Albaugh]

Eric Meijer (tgakem@pebbles.chem.tue.nl) wrote:

: [...], and so on, so if the name mangling were made the same, your
: programs would link against libraries provided from other compilers
: but then crash when run. For this reason, the ARM encourages compiler
: writers to make their name mangling different from that of other
: compilers for the same platform. Incompatible libraries are then
: detected at link time, rather than at run time.

: ------------------------------------------------------------------------

: This means that you need to compile all C++ source in an application
: with one and the same compiler.  With C libraries this problem generally
: doesn't exist.

	Beg to differ. On _many_ platforms, the C libraries from
different vendors (and even from different switch settings to the
same compiler) can and do differ in such things as parameter types,
structure layouts, etc. Heck MSVC's "intrinsic" functions disagree
with their own .h files, sometimes :-) Anyway, the result of these
"bleeding ends" is not typically detected at link time, but either
at compile time (_if_ the programmer cranks warnings up high enough
_and_ the .h files are complete and correct) or at run-time, where
they manifest themselves as obscure behavior.

						Mike
| albaugh@agames.com, speaking only for myself

Re: Linking problems, mangling ?

[Roland TTK]

Eric Meijer a dit:
 >You are trying to do something which cannot work.
Yuk !

 >This means that you need to compile all C++ source in an application with
 >one and the same compiler.  With C libraries this problem generally doesn't
 >exist.
Thanks a lot for your help. But ...
I have to use these libraries and I don't have the source code so I can't
recompile them with gcc. So the solution should be to compile my sources with
the sun C++ compiler. But this one doesn't have the STL. So I tried to compile
the libstdc++ in order to emulate the STL.
Alas the sun compiler was unable to do it because libstdc++ sources contain
too many gcc specific things.

So I'm f---ed  :(

Or maybe I could find another implementation of the STL ??

Thanks.
-- 
 TTK
 http://www.multimania.com/ttk22

Linking problems, mangling ?

[Roland TTK]

 Hi
I have (succesfully ?) compiled my sources under solaris 2.6 with g++ 2.95 but
I get undefined reference errors when linking. I try to link my object files
with a library which was compiled by the sun C++ compiler. It seems like this
compiler uses a specific mangling grammar, so the linker can't find the
definitions it needs.

I think the mangling grammar is pretty old (1994) so I can't beleive gcc
doesn't emulate it.

Help !

Thanks for your inputs.

-- 
 TTK
 http://www.multimania.com/ttk22