On Fri, 2013-12-27 at 11:58 +0100, Torsten Meißner wrote:
> with the GNAT_LARGS argument for the Makefile and -gnatwa I can build until 
> the linker starts its work:
> 
> 
> gnatlink ghdl_mcode.ali -g memsegs_c.o chkstk.o -Wl,-no_compact_unwind i386.o 
> linux.o times.o grt-cbinding.o grt-cvpi.o
> ld: warning: can't find atom for N_GSYM stabs 
> ada_main__gnat_version:G(0,12)=k(0,13)=B(0,14)=ar(0,9);1;67;(0,4) in 
> b~ghdl_mcode.o
> ld: warning: can't find atom for N_GSYM stabs 
> ada_main__ada_main_program_name:G(0,15)=k(0,16)=B(0,17)=ar(0,9);1;16;(0,4) in 
> b~ghdl_mcode.o
> ld: warning: can't find atom for N_GSYM stabs 
> ada_main__u00001:G(0,18)=k(0,19)=B(0,20)=r(0,20);0;037777777777; in 
> b~ghdl_mcode.o

> Don’t know if this is an error, which prevents ghdl from working… 


I think it is...

it looks like something missing in the Ada standard library...

> GNA dir bug18351:
> analyze 18351.vhd
> elaborate and simulate problem
> ../../testenv.sh: line 79:  4004 Segmentation fault: 11  $GHDL --elab-run 
> $GHDL_FLAGS $GHDL_ELABFLAGS $@
which manifests itself when you run the executable.

> I build with the GNAT tools, not with Xcode:
> which gcc
> /usr/local/ada-4.3/bin/gcc

It has been observed before that this is a rather old Gnat installation.
ghdl has been successfully built using FSF Gnat 4.6.3, 4.7.2,
4.8.various, Gnat GPL 2013 and some other versions.

David's (successful) builds use Gnat GPL 2013 as the Ada compiler; I
don't know offhand which compilers he gave up on...

Some builds (mine at least) have been bootstrap builds, also building
libada from the same source code release.

I suspect you may be compiling a 4.8 compiler and attempting to link
against a 4.3 libada/runtime system : if so, problems do not surprise
me.

Is there a compelling reason you need to build with 4.3? If so, you will
at least need to configure with --enable-libada and verify that it is
building and installing correctly; IMO it would be simpler to upgrade
the Ada compiler instead.

Thanks for persisting and reporting : knowing what DOESN'T work is at
least as important as knowing what does...

- Brian


_______________________________________________
Ghdl-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/ghdl-discuss

Reply via email to