Re: [PATCH] slab.h: Avoid using & for logical and of booleans

From: Alexander Duyck <alexander.duyck@gmail.com>
To: bvanassche@acm.org
Cc: linux@rasmusvillemoes.dk,
	Andrew Morton <akpm@linux-foundation.org>,
	LKML <linux-kernel@vger.kernel.org>,
	Vlastimil Babka <vbabka@suse.cz>,
	Mel Gorman <mgorman@techsingularity.net>,
	Christoph Lameter <cl@linux.com>,
	guro@fb.com, Pekka Enberg <penberg@kernel.org>,
	David Rientjes <rientjes@google.com>,
	Joonsoo Kim <iamjoonsoo.kim@lge.com>,
	linux-mm <linux-mm@kvack.org>
Subject: Re: [PATCH] slab.h: Avoid using & for logical and of booleans
Date: Tue, 6 Nov 2018 10:17:36 -0800	[thread overview]
Message-ID: <CAKgT0UfbLH07Dqgk0HnhcoPbmQnoykbXLYzkdqTXkafh952Ceg@mail.gmail.com> (raw)
In-Reply-To: <1541526521.196084.184.camel@acm.org>

On Tue, Nov 6, 2018 at 9:48 AM Bart Van Assche <bvanassche@acm.org> wrote:
>
> On Tue, 2018-11-06 at 09:20 -0800, Alexander Duyck wrote:
> > On Mon, Nov 5, 2018 at 4:32 PM Bart Van Assche <bvanassche@acm.org> wrote:
> > >
> > > On Mon, 2018-11-05 at 16:11 -0800, Alexander Duyck wrote:
> > > > If we really don't care then why even bother with the switch statement
> > > > anyway? It seems like you could just do one ternary operator and be
> > > > done with it. Basically all you need is:
> > > > return (defined(CONFIG_ZONE_DMA) && (flags & __GFP_DMA)) ? KMALLOC_DMA :
> > > >         (flags & __GFP_RECLAIMABLE) ? KMALLOC_RECLAIM : 0;
> > > >
> > > > Why bother with all the extra complexity of the switch statement?
> > >
> > > I don't think that defined() can be used in a C expression. Hence the
> > > IS_ENABLED() macro. If you fix that, leave out four superfluous parentheses,
> > > test your patch, post that patch and cc me then I will add my Reviewed-by.
> >
> > Actually the defined macro is used multiple spots in if statements
> > throughout the kernel.
>
> The only 'if (defined(' matches I found in the kernel tree that are not
> preprocessor statements occur in Perl code. Maybe I overlooked something?

You may be right. I think I was thinking of "__is_defined", not "defined".

> > The reason for IS_ENABLED is to address the fact that we can be
> > dealing with macros that indicate if they are built in or a module
> > since those end up being two different defines depending on if you
> > select 'y' or 'm'.
>
> From Documentation/process/coding-style.rst:
>
> Within code, where possible, use the IS_ENABLED macro to convert a Kconfig
> symbol into a C boolean expression, and use it in a normal C conditional:
>
> .. code-block:: c
>
>         if (IS_ENABLED(CONFIG_SOMETHING)) {
>                 ...
>         }
>
> Bart.

Right. Part of the reason for suggesting that is that depending on how
you define "CONFIG_SOMETHING" it can actually be defined as
"CONFIG_SOMETHING" or "CONFIG_SOMETHING_MODULE".  I was operating
under the assumption that CONFIG_ZONE_DMA wasn't ever going to be
built as a module.

Thanks.

- Alex