[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CACRpkdYAjj+EuF+iu4fKjt2Cviu8V+U66HnQThawwU58UGRUzQ@mail.gmail.com>
Date: Fri, 4 Oct 2019 21:41:12 +0200
From: Linus Walleij <linus.walleij@...aro.org>
To: Marco Felsch <m.felsch@...gutronix.de>
Cc: Support Opensource <support.opensource@...semi.com>,
Lee Jones <lee.jones@...aro.org>,
Rob Herring <robh+dt@...nel.org>,
Liam Girdwood <lgirdwood@...il.com>,
Mark Brown <broonie@...nel.org>, stwiss.opensource@...semi.com,
Sascha Hauer <kernel@...gutronix.de>,
"open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS"
<devicetree@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 3/5] regulator: da9062: add voltage selection gpio support
On Tue, Sep 17, 2019 at 2:43 PM Marco Felsch <m.felsch@...gutronix.de> wrote:
> + /*
> + * We only must ensure that the gpio device is probed before the
> + * regulator driver so no need to store the reference global. Luckily
> + * devm_* releases the gpio upon a unbound action.
> + */
> + gpi = devm_gpiod_get_from_of_node(cfg->dev, np, prop, 0, GPIOD_IN |
> + GPIOD_FLAGS_BIT_NONEXCLUSIVE, label);
Do you really need the GPIOD_FLAGS_BIT_NONEXCLUSIVE flag here?
I don't think so, but describe what usecase you have that warrants this
being claimed twice. Normally that is just needed when you let the
regulator core handle enablement of a regulator over GPIO, i.e.
ena_gpiod in struct regulator_config.
> + /* We need the local number */
> + nr = da9062_gpio_get_hwgpio(gpi);
If you really need this we should add a public API to gpiolib and not
create custom APIs.
Just make a patch adding
int gpiod_to_offset(struct gpio_desc *d);
to the public gpiolib API in include/linux/gpio/consumer.h
and add the code in gpiolib.c to do this trick.
Yours,
Linus Walleij
Powered by blists - more mailing lists