[RFC] unify dynamic_symbol_p implementations

[Alan Modra]

On Sun, Jul 20, 2003 at 08:26:43AM -0700, H. J. Lu wrote:
> Weak defined symbols are purely for link time and only meaningful for
> relocatable inputs.

Hmm, OK.  Seems a little odd that we make weak undef dynamic then.

I'm going to install my patch so that x86 and ppc work again.  It's
likely that ia64 could use _bfd_elf_symbol_refs_local_p instead of
_bfd_elf_dynamic_symbol_p, but I don't really have the time right
now to figure out what needs doing.

-- 
Alan Modra
IBM OzLabs - Linux Technology Centre