Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20191127135932.7223-1-m.felsch@pengutronix.de> <20191127135932.7223-2-m.felsch@pengutronix.de>
In-Reply-To: <20191127135932.7223-2-m.felsch@pengutronix.de>
From:   Linus Walleij <linus.walleij@linaro.org>
Date:   Fri, 29 Nov 2019 10:32:51 +0100
Message-ID: <CACRpkdbG=XiQHNZa+zBqdyTDRhyXD5rLxbLjp3qqGbcQeTX26Q@mail.gmail.com>
Subject: Re: [PATCH v2 1/5] gpio: add support to get local gpio number
To:     Marco Felsch <m.felsch@pengutronix.de>
Cc:     Bartosz Golaszewski <bgolaszewski@baylibre.com>,
        Support Opensource <support.opensource@diasemi.com>,
        Lee Jones <lee.jones@linaro.org>,
        Rob Herring <robh+dt@kernel.org>,
        Liam Girdwood <lgirdwood@gmail.com>,
        Mark Brown <broonie@kernel.org>, stwiss.opensource@diasemi.com,
        Adam Thomson <Adam.Thomson.Opensource@diasemi.com>,
        "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" 
        <devicetree@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Sascha Hauer <kernel@pengutronix.de>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Marco,

thanks for your patch!

On Wed, Nov 27, 2019 at 2:59 PM Marco Felsch <m.felsch@pengutronix.de> wrote:

> Sometimes consumers needs to know the gpio-chip local gpio number of a
> 'struct gpio_desc' for further configuration. This is often the case for
> mfd devices.
>
> Signed-off-by: Marco Felsch <m.felsch@pengutronix.de>
(...)
> +int gpiod_to_offset(struct gpio_desc *desc)
> +{
> +       return gpio_chip_hwgpio(desc);
> +}
> +EXPORT_SYMBOL_GPL(gpiod_to_offset);

That seems like an unnecessary wrapper.

What about renaming gpio_chip_hwgpio() everywhere
to gpiod_to_offet(), remove it from drivers/gpio/gpiolib.h
and export it in <linux/gpio/consumer.h> instead?

I suppose this is what Bartosz is indicating, not sure though.

Indeed it is a bit of a worrysome thing to export and we need
to be very specific about its usecase, so I'd also like some
nice to-the-point kerneldoc on the export site so that it is
clear what corner cases this function is for. (Like in this
specific driver.)

Yours,
Linus Walleij