Re: Port to x86-64

[Andreas Jaeger <aj at suse dot de>]

Andrew Cagney <ac131313@cygnus.com> writes:

> Jiri Smid wrote:
> > 
> >   Hi,
> > 
> > Andreas Jaeger <aj@suse.de> writes:
> > 
> > > >
> > > >     o       submit the configury mess early
> > > >
> > > >             even if you don't have real code
> > > >             could I suggest thinking about
> > > >             separating this out so that
> > > >             you can get on with the real
> > > >             work.
> > > >
> > 
> >   Before I will submit the configury mess I would like to little discuss
> > about it.
> >   The generic name of CPU is x86_64 in already ported software. It seems
> > that there shoud be create new config directory gdb/config/x86_64. But
> > this processor is extended i386 architecture and many parts of code will
> > be the same as in i386.
> >   I think that the best solution is to create this new config directory
> > x86_64 and from tm-x86_64.h include i386/tm-i386.h and redefine some
> > declarations. What do you think?
> 
> What convention, if any, was adopted by GCC and what do config.guess /
> config.sub return?

config.guess/config.sub return x86_64 since x86-64 would cause
problems since the "-" is special for config.

> I suspect that ``x86-64'' is going to be more acceptable than x86_64 a
> file name.

As filename, definitly.  For GCC the support is part of the i386
file.  For binutils, the assembler/dissassembler is shared with i386,
but we have also:
./bfd/elf64-x86-64.c
./include/elf/x86-64.h

Andreas
-- 
 Andreas Jaeger
  SuSE Labs aj@suse.de
   private aj@arthur.inka.de
    http://www.suse.de/~aj