[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CACRpkdbNvjELQ_9bJSmD5oqhv+uHhSwBErGcKsZayCntDbTwPA@mail.gmail.com>
Date: Thu, 21 Feb 2019 11:44:26 +0100
From: Linus Walleij <linus.walleij@...aro.org>
To: Shravan Kumar Ramani <sramani@...lanox.com>
Cc: Bartosz Golaszewski <bgolaszewski@...libre.com>,
"open list:GPIO SUBSYSTEM" <linux-gpio@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3 1/1] gpio: add driver for Mellanox BlueField GPIO controller
Hi Shravan,
thank you for your patch!
On Wed, Feb 20, 2019 at 11:07 PM Shravan Kumar Ramani
<sramani@...lanox.com> wrote:
> This patch adds support for the GPIO controller used by Mellanox
> BlueField SOCs.
>
> Reviewed-by: David Woods <dwoods@...lanox.com>
> Signed-off-by: Shravan Kumar Ramani <sramani@...lanox.com>
(...)
> +config GPIO_MLXBF
> + tristate "Mellanox BlueField SoC GPIO"
> + depends on (MELLANOX_PLATFORM && ARM64 && ACPI) || COMPILE_TEST
Add
select GPIO_GENERIC
Becaus I think you can use it.
> +static int mlxbf_gpio_set_input(struct gpio_chip *chip, unsigned int offset)
> +{
> + struct mlxbf_gpio_state *gs = gpiochip_get_data(chip);
> + u64 in;
> + u64 out;
> +
> + out = readq(gs->dc_base + MLXBF_GPIO_PIN_DIR_O);
> + in = readq(gs->dc_base + MLXBF_GPIO_PIN_DIR_I);
> +
> + spin_lock(&gs->lock);
> + writeq(out & ~BIT(offset), gs->dc_base + MLXBF_GPIO_PIN_DIR_O);
> + writeq(in | BIT(offset), gs->dc_base + MLXBF_GPIO_PIN_DIR_I);
> + spin_unlock(&gs->lock);
> +
> + return 0;
> +}
> +
> +static int mlxbf_gpio_set_output(struct gpio_chip *chip, unsigned int offset,
> + int value)
> +{
> + struct mlxbf_gpio_state *gs = gpiochip_get_data(chip);
> + u64 in;
> + u64 out;
> +
> + out = readq(gs->dc_base + MLXBF_GPIO_PIN_DIR_O);
> + in = readq(gs->dc_base + MLXBF_GPIO_PIN_DIR_I);
> +
> + spin_lock(&gs->lock);
> + writeq(out | BIT(offset), gs->dc_base + MLXBF_GPIO_PIN_DIR_O);
> + writeq(in & ~BIT(offset), gs->gs->dc_base + MLXBF_GPIO_PIN_DIR_Idc_base + MLXBF_GPIO_PIN_DIR_I);
> + spin_unlock(&gs->lock);
> +
> + return 0;
> +}
> +
> +static int mlxbf_gpio_get(struct gpio_chip *chip, unsigned int offset)
> +{
> + u64 value;
> + struct mlxbf_gpio_state *gs = gpiochip_get_data(chip);
> +
> + spin_lock(&gs->lock);
> + value = readq(gs->dc_base + MLXBF_GPIO_PIN_STATE);
> + spin_unlock(&gs->lock);
> +
> + return (value >> offset) & 1;
> +}
> +
> +static void mlxbf_gpio_set(struct gpio_chip *chip, unsigned int offset,
> + int value)
> +{
> + u64 data;
> + struct mlxbf_gpio_state *gs = gpiochip_get_data(chip);
> +
> + spin_lock(&gs->lock);
> + data = readq(gs->dc_base + MLXBF_GPIO_PIN_STATE);
> +
> + if (value)
> + data |= BIT(offset);
> + else
> + data &= ~BIT(offset);
> + writeq(data, gs->dc_base + MLXBF_GPIO_PIN_STATE);
> + spin_unlock(&gs->lock);
> +}
This looks like it can use the generic MMIO library.
Look at other drivers calling bgpio_init() to set up set/get/direction
helpers for inspiration. The MMIO library should be able to deal with
64bit registers IIUC. With this approach you get get/set_multiple
for free.
There is very detailed documenttion above the function bgpio_init()
in drivers/gpio/gpio-mmio.c.
I think something like this in probe():
ret = bgpio_init(gc, dev, 8,
gs->dc_base + MLXBF_GPIO_PIN_STATE,
NULL,
NULL,
gs->dc_base + MLXBF_GPIO_PIN_DIR_O,
gs->dc_base + MLXBF_GPIO_PIN_DIR_I,
0);
if (ret)
return -ENODEV;
gc->label = dev_name(dev);
gc->parent = &pdev->dev;
gc->owner = THIS_MODULE;
gc->base = -1;
gc->ngpio = MLXBF_GPIO_NR;
(...)
This makes the driver short and efficient and reuse the MMIO library
in a nice way. It also implements the spinlock for you so you don't
need that anymore.
Yours,
Linus Walleij
Powered by blists - more mailing lists