Re: Linker error on unstable amd64

[Kurt Roeckx <kurt@roeckx.be>]

On Tue, Jan 23, 2007 at 10:35:32PM +0100, Goswin von Brederlow wrote:
> > ld -r -o goto-programs.o goto_convert.o goto_function.o goto_main.o goto_sideeffects.o goto_program.o basic_blocks.o goto_threads.o goto_check.o goto_function_pointers.o goto_functions.o goto_inline.o remove_skip.o function_dependencies.o goto_convert_functions.o builtin_functions.o show_claims.o destructor.o set_claims.o slicer.o invariant_set.o invariant_propagation.o add_race_assertions.o rw_set.o read_goto_binary.o read_goto_object.o xml_goto_function.o xml_goto_program.o xml_goto_program_hashing.o xml_goto_function_hashing.o invariant_set_domain.o static_analysis.o
> 
> That is the _C_ linker.

The same linker is used for both C and C++.

However, the weird part here is that it's creating a relocatable object
(-r switch, .o file) from other relocatable objects.  I've atleast seen
weird behaviour that way and it's probably better to just create an
archive out of those .o files.  I'm not at all sure if that's going to
work, it'll probably just delay the problem until it gets really linked.
And this way is also supposed to work.


Kurt