Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
MIME-Version: 1.0
References: <20220426131102.23966-1-andrea.merello@gmail.com>
 <20220426131102.23966-9-andrea.merello@gmail.com> <CAHp75VdLiBkg100UjFN36rW_vaOObOoJ_Mv9n=4LjSWb+dQWMw@mail.gmail.com>
 <CAN8YU5PzwmeQ9XA3qod7HejG6cCLCrPvda5eomCh5hUze_DWcA@mail.gmail.com> <CAHp75VfOZpD135q_eERnLk0NorXwPxY8DFbKMu+eKV8XahGC1A@mail.gmail.com>
In-Reply-To: <CAHp75VfOZpD135q_eERnLk0NorXwPxY8DFbKMu+eKV8XahGC1A@mail.gmail.com>
Reply-To: andrea.merello@gmail.com
From:   Andrea Merello <andrea.merello@gmail.com>
Date:   Mon, 2 May 2022 15:12:45 +0200
Message-ID: <CAN8YU5P_vGV0vrhdy273ef9GH2Y2=TbXbsiho=V9vG44NXAujA@mail.gmail.com>
Subject: Re: [v5 08/14] iio: imu: add Bosch Sensortec BNO055 core driver
To:     Andy Shevchenko <andy.shevchenko@gmail.com>
Cc:     Jonathan Cameron <jic23@kernel.org>,
        Mauro Carvalho Chehab <mchehab+huawei@kernel.org>,
        linux-iio <linux-iio@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        devicetree <devicetree@vger.kernel.org>,
        Lars-Peter Clausen <lars@metafoo.de>,
        Rob Herring <robh+dt@kernel.org>,
        Matt Ranostay <matt.ranostay@konsulko.com>,
        Alexandru Ardelean <ardeleanalex@gmail.com>,
        jmondi <jacopo@jmondi.org>,
        Andrea Merello <andrea.merello@iit.it>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Il giorno lun 2 mag 2022 alle ore 12:12 Andy Shevchenko
<andy.shevchenko@gmail.com> ha scritto:

One inline comment. OK for the rest

> > > > +#define BNO055_ATTR_VALS(...)          \
> > > > +       .vals = (int[]){ __VA_ARGS__},  \
> > > > +       .len = ARRAY_SIZE(((int[]){__VA_ARGS__}))

[...]

> And my point about readability. The reader, and even the author after
> some time, may have no clue in this forest of the macros and castings
> what's going on.

While I'm OK wrt your point in general, consider that it's just a
three LOC macro, used only in a few structs just below. I wouldn't say
it's so inricated; I've seen by far worse in the kernel :)

> > but about avoiding as much as
> > possible bugs caused by mismatched attr_vals, attr_aux and
> > ARRAY_SIZE() arg. e.g:
> > bno055_sysfs_attr_avail(priv, bno_foo_vals, ARRAY_SIZE(bno_bar_vals),
> > bno_foobar_aux, vals, len)
> >
> > I used to make quite a lot of mess until I grouped all the stuff in
> > one struct :/
>
> If something you want to prevent at compile time, consider to utilize
> static_assert() and / or BUILD_BUG_ON() depending on the place in the
> code (the former is preferred).

I would be happy to get rid of my macro and use those assertion
things, but I can't see how exactly. Do you have any advice about how
to take advantage of them for catching bugs like the one above in this
specific case?