Re: [patch]: Merge from gnustep for obj-c. Remove warnings for target w64.

[Andrew Thomas Pinski <pinskia at gmail dot com>]

Sent from my iPhone

On Nov 19, 2008, at 12:19 AM, Kai Tietz <Kai.Tietz@onevision.com> wrote:

> "Danny Smith" <dansmister@gmail.com> wrote on 19.11.2008 03:58:14:
>
> > >       * objc/objc.h (BOOL): Prevent redeclaration of BOOL, if  
> > > already
> > > there.
> >
> >
> > --- gcc.orig/libobjc/objc/objc.h
> > +++ gcc/libobjc/objc/objc.h
> > @@ -39,8 +39,10 @@ extern "C" {
> > #ifdef __vxworks
> > typedef int BOOL;
> > #else
> > +#ifndef BOOL
> > typedef unsigned char  BOOL;
> > #endif
> > +#endif
> > #define YES   (BOOL)1
> > #define NO    (BOOL)0
> >
> >
> > That doe not seem right.  It means that if w32api's windef.h is
> > included before the objc.h language header, then BOOL may be
> > equivalent to WINBOOL which is an int.  If if is included after then
> > is in an unsigned char.   Wouldn't it be better to just  #iundef BOOL
> > before the typedef?
> >
> > Danny
>
> Well, I see in both cases a problem here. So to have the pattern
> #ifdef BOOL
> #undef BOOL
> #endif
> typedef unsigned char BOOL;

You can use #undef unconditional so no need for the ifdef.
This part of the patch is ok but I have to look into the other parts  
still.


> and the pattern
> #ifndef BOOL
> typedef unsigned char BOOL;
> #endif
>
> are equivalent IMHO. But I open-minded for this, but see no real  
> advantage
> between one of the variants.
>
> Kai
>
> |  (\_/)  This is Bunny. Copy and paste Bunny
> | (='.'=) into your signature to help him gain
> | (")_(") world domination.