Received-SPF: pass (google.com: domain of linux-kernel+bounces-30928-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Precedence: bulk
MIME-Version: 1.0
References: <20240118-ktd2801-v2-0-425cf32e0769@skole.hr> <20240118-ktd2801-v2-2-425cf32e0769@skole.hr>
In-Reply-To: <20240118-ktd2801-v2-2-425cf32e0769@skole.hr>
From: Linus Walleij <linus.walleij@linaro.org>
Date: Fri, 19 Jan 2024 10:02:33 +0100
Message-ID: <CACRpkdaUvmmbGUyQ-L_u8c73=Oz+qE88GXd1=cUY7r+PPttJbw@mail.gmail.com>
Subject: Re: [PATCH v2 2/2] backlight: Add Kinetic KTD2801 driver
To: =?UTF-8?Q?Duje_Mihanovi=C4=87?= <duje.mihanovic@skole.hr>
Cc: Lee Jones <lee@kernel.org>, Daniel Thompson <daniel.thompson@linaro.org>, 
	Jingoo Han <jingoohan1@gmail.com>, Pavel Machek <pavel@ucw.cz>, Rob Herring <robh+dt@kernel.org>, 
	Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>, Conor Dooley <conor+dt@kernel.org>, 
	Helge Deller <deller@gmx.de>, Karel Balej <balejk@matfyz.cz>, ~postmarketos/upstreaming@lists.sr.ht, 
	dri-devel@lists.freedesktop.org, linux-leds@vger.kernel.org, 
	devicetree@vger.kernel.org, linux-kernel@vger.kernel.org, 
	linux-fbdev@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hi Duje,

thanks for your patch!

On Thu, Jan 18, 2024 at 6:33=E2=80=AFPM Duje Mihanovi=C4=87 <duje.mihanovic=
@skole.hr> wrote:

> Add driver for the Kinetic KTD2801 backlight driver.>
> Signed-off-by: Duje Mihanovi=C4=87 <duje.mihanovic@skole.hr>

Add some commit message?

> +#include <linux/backlight.h>
> +#include <linux/delay.h>
> +#include <linux/gpio/consumer.h>
> +#include <linux/of.h>

I don't think you need <linux/of.h>, the compatible table works without
that (is in the device driver core).

> +/* These values have been extracted from Samsung's driver. */
> +#define KTD2801_EXPRESSWIRE_DETECT_DELAY_US    150
> +#define KTD2801_EXPRESSWIRE_DETECT_US          270
> +#define KTD2801_LOW_BIT_HIGH_TIME_US           5
> +#define KTD2801_LOW_BIT_LOW_TIME_US            (4 * KTD2801_HIGH_BIT_LOW=
_TIME_US)
> +#define KTD2801_HIGH_BIT_LOW_TIME_US           5
> +#define KTD2801_HIGH_BIT_HIGH_TIME_US          (4 * KTD2801_HIGH_BIT_LOW=
_TIME_US)
> +#define KTD2801_DATA_START_US                  5
> +#define KTD2801_END_OF_DATA_LOW_US             10
> +#define KTD2801_END_OF_DATA_HIGH_US            350
> +#define KTD2801_PWR_DOWN_DELAY_US              2600
> +
> +#define KTD2801_DEFAULT_BRIGHTNESS     100
> +#define KTD2801_MAX_BRIGHTNESS         255
> +
> +struct ktd2801_backlight {
> +       struct backlight_device *bd;
> +       struct gpio_desc *gpiod;
> +       bool was_on;
> +};
> +
> +static int ktd2801_update_status(struct backlight_device *bd)
> +{
> +       struct ktd2801_backlight *ktd2801 =3D bl_get_data(bd);
> +       u8 brightness =3D (u8) backlight_get_brightness(bd);
> +
> +       if (backlight_is_blank(bd)) {
> +               gpiod_set_value(ktd2801->gpiod, 0);
> +               udelay(KTD2801_PWR_DOWN_DELAY_US);

That's 2600 us, a pretty long delay in a hard loop or delay timer!

Can you use usleep_range() instead, at least for this one?

> +       for (int i =3D 0; i < 8; i++) {
> +               u8 next_bit =3D (brightness & 0x80) >> 7;

I would just:

#include <linux/bits.h>

bool next_bit =3D !!(brightness & BIT(7));

Yours,
Linus Walleij