Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20200617220226.88948-1-luc.vanoostenryck@gmail.com>
 <CANiq72nNQ+11ch7sRjLLc33s_OkftuOqgGkCDggq88tbJtdPNQ@mail.gmail.com> <20200618050249.lqwznznltm3lcykr@ltop.local>
In-Reply-To: <20200618050249.lqwznznltm3lcykr@ltop.local>
From:   Geert Uytterhoeven <geert@linux-m68k.org>
Date:   Fri, 19 Jun 2020 10:07:34 +0200
Message-ID: <CAMuHMdWQsirja-h3wBcZezk+H2Q_HShhAks8Hc8ps5fTAp=ObQ@mail.gmail.com>
Subject: Re: [PATCH] sparse: use identifiers to define address spaces
To:     Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Cc:     Miguel Ojeda <miguel.ojeda.sandonis@gmail.com>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        linux-kernel <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Luc,

On Thu, Jun 18, 2020 at 7:05 AM Luc Van Oostenryck
<luc.vanoostenryck@gmail.com> wrote:
> On Thu, Jun 18, 2020 at 03:22:15AM +0200, Miguel Ojeda wrote:
> > On Thu, Jun 18, 2020 at 12:02 AM Luc Van Oostenryck
> > <luc.vanoostenryck@gmail.com> wrote:
> > >
> > > diff --git a/include/linux/compiler_types.h b/include/linux/compiler_types.h
> > > index 21aed0981edf..e368384445b6 100644
> > > --- a/include/linux/compiler_types.h
> > > +++ b/include/linux/compiler_types.h
> > > @@ -5,20 +5,20 @@
> > >  #ifndef __ASSEMBLY__
> > >
> > >  #ifdef __CHECKER__
> > > -# define __user                __attribute__((noderef, address_space(1)))
> > >  # define __kernel      __attribute__((address_space(0)))
> > > +# define __user                __attribute__((noderef, address_space(__user)))
> >
> > I guess `__kernel` moves to the first place since it uses the first
> > address space?
>
> No, there is no really an order between address spaces. Even before
> this patch, the number were only used as an ID to distinguish them
> from each other.
>
> I just moved __kernel above because it is quite different from
> the others because it's the default one, and so:
> * it's never displayed
> * it's normally not needed, nor in type annotations, nor in cast
>   between address spaces. The only time it's needed is when it's
>   combined with a typeof to express "the same type as this one but
>   without the address space"
> * it can't be defined with a name, '0' must be used.
>
> So, it seemed strange to me to have it in the middle of the other ones.

Indeed. It looks like this whole list is completely unsorted, and was created
by appending new definitions at the bottom.
So perhaps all the address_space ones should be grouped together?

Gr{oetje,eeting}s,

                        Geert


--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds