Re: [drow@mvista.com: Re: RFA: LOC_COMPUTED support]

[Joel Brobecker <brobecker@gnat.com>]

> > I think I found some switches that still need LOC_COMPUTED{,_ARG} cases:
> > - ada-lang.c (ada_resolve_subexp, symtab_for_sym,
> >     ada_add_block_symbols, fill_in_ada_prototype) 
> 
> OK, what's up with that?  The code bits in ada_resolve_subexp all
> appear to be commented out.  Is this code going away?

Just from the top of my head, since I haven't directly been involved
with the process of getting this file in: I think they are commented
out until we can add the different pieces that would make them
compilable and/or activatable. 

They will for sure be re-examined when we activate this code.

> symtab_for_sym is pretty hokey too.  All the Ada code reads like
> someone wanted to keep changes local to ada-lang.c when they were
> addressing general symtab weaknesses.  I hope it can all go away
> someday.  The amount of duplication is pretty scary, too.

I don't know the history of this code, Paul Hilfinger will probably know
more, but I agree that getting rid of this code one day would be a
tremendous improvement.

Thanks for taking care of ada-lang.c in the interim. If any problem
arises with the maintenance of these file, please let me and Paul know,
we'll jump in to help.

-- 
Joel