Subject: Re: stdbool.h compatibility with gcc 2.x
To: None <tech-userlevel@netbsd.org>
From: Julio M. Merino Vidal <jmmv84@gmail.com>
List: tech-userlevel
Date: 09/30/2005 16:58:35
On 9/30/05, James Chacon <jmc@netbsd.org> wrote:
> On Fri, Sep 30, 2005 at 04:51:54PM +0200, Julio M. Merino Vidal wrote:
> > Hi all,
> >
> > I've been noticed that we can't use stdbool.h in programs that form par=
t
> > of NetBSD because it is not provided by all ports (e.g., vax, which sti=
ll
> > has an old gcc), thus breaking the build.
> >
> > I believe that having a boolean type available (and usable) could make
> > our code clearer in some circumstances.
> >
> > So, I'm wondering... could it be silly to provide a "fake" stdbool.h he=
ader
> > that uses an integer type to define the boolean type in those platforms
> > that don't have it (i.e., HAVE_GCC3=3Dno)?  Should be trivial to do.
> >
>
> There's only 1 place in-tree right now I think that required stdbool.h
> and it was specifically patched for vax already.
>
> What else do we have that requires/needs stdbool now vs waiting for the
> next gcc update which will address vax?

Nothing I'm aware of (after the fix I just commited).  But it has happened =
to
me twice (wsmoused, mount_tmpfs) to add some code that later needs to
be "fixed" because it used stdbool.h.

--
Julio M. Merino Vidal <jmmv84@gmail.com>
http://www.livejournal.com/users/jmmv/
The NetBSD Project - http://www.NetBSD.org/