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;
Date:   Wed, 14 Jul 2021 18:54:08 +0300
From:   Alexey Dobriyan <adobriyan@gmail.com>
To:     Christoph Hellwig <hch@infradead.org>
Cc:     akpm@linux-foundation.org, linux-kbuild@vger.kernel.org,
        linux-kernel@vger.kernel.org, linux-arch@vger.kernel.org,
        arnd@arndb.de
Subject: Re: [PATCH] Decouple build from userspace headers
Message-ID: <YO8IoNwXS4h26+9v@localhost.localdomain>
References: <YO3txvw87MjKfdpq@localhost.localdomain>
 <YO7zEFNSXOY8pKCQ@infradead.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <YO7zEFNSXOY8pKCQ@infradead.org>
Precedence: bulk

On Wed, Jul 14, 2021 at 03:22:08PM +0100, Christoph Hellwig wrote:
> > -#define signals_blocked false
> > +#define signals_blocked 0
> 
> Why can't we get at the kernel definition of false here?

Variable and other code surrounding this wants "int".
I don't really want to expand into bool conversion.

> > new file mode 100644
> > --- /dev/null
> > +++ b/include/stdarg.h
> > @@ -0,0 +1,9 @@
> > +#ifndef _LINUX_STDARG_H
> > +#define _LINUX_STDARG_H
> > +typedef __builtin_va_list __gnuc_va_list;
> > +typedef __builtin_va_list va_list;
> > +#define va_start(v, l)	__builtin_va_start(v, l)
> > +#define va_end(v)	__builtin_va_end(v)
> > +#define va_arg(v, T)	__builtin_va_arg(v, T)
> > +#define va_copy(d, s)	__builtin_va_copy(d, s)
> > +#endif
> 
> Empty lines before and after the include guards would be nice.
> 
> What do we need the __gnuc_va_list typedef for?

That's because without __gnuc_va_list something didn't compile.
I'm preparing second version with <linux/stdarg.h> where __gnuc_va_list is
unnecessary indeed.

> Otherwise this looks great.  As a follow on maybe move the new header
> to <linux/stdarg.h> to make clear to everyone that we are using our
> own version.