[v4,03/10] pwm: max7360: Add MAX7360 PWM support

From: Kamel Bouhara <kamel.bouhara@bootlin.com>

Add driver for Maxim Integrated MAX7360 PWM controller, supporting up to
8 independent PWM outputs.

Signed-off-by: Kamel Bouhara <kamel.bouhara@bootlin.com>
Co-developed-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
---
 drivers/pwm/Kconfig       |  10 +++
 drivers/pwm/Makefile      |   1 +
 drivers/pwm/pwm-max7360.c | 213 ++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 224 insertions(+)

On Fri, Feb 14, 2025 at 12:49:53PM +0100, mathieu.dubois-briand@bootlin.com wrote:
> From: Kamel Bouhara <kamel.bouhara@bootlin.com>
> 
> Add driver for Maxim Integrated MAX7360 PWM controller, supporting up to
> 8 independent PWM outputs.

...

+ bits.h

+ dev_printk.h

> +#include <linux/err.h>


> +#include <linux/math.h>

Other way around, id est you need math64.h (see below).

> +#include <linux/mfd/max7360.h>

+ minmax.h

> +#include <linux/mod_devicetable.h>
> +#include <linux/module.h>

> +#include <linux/of.h>

Is this used? Cargo cult?

> +#include <linux/platform_device.h>
> +#include <linux/pwm.h>
> +#include <linux/regmap.h>

+ types.h

...

> +#define MAX7360_PWM_PERIOD_NS			2000000 /* 500 Hz */

Comment is superfluous, if you need HZ units, define the respective one.
Also you can use something like (2 * NSEC_PER_MSEC) which will immediately
gives a hint of how long this is and reduces potential 0:s miscalculations.
This will need time.h

...

> +#define MAX7360_PWM_CTRL_ENABLE(n)		BIT(n)
> +#define MAX7360_PWM_PORT(n)			BIT(n)

Personally I find these macros overkill. The value of them much shorter and
equally readable.

...

> +struct max7360_pwm {

> +	struct device *parent;

Is it not the same as you can derive from regmap?

> +	struct regmap *regmap;

Btw, have you checked the code generation if you place regmap the first in the
structure? It might affect it.

> +};

...

> +	/*
> +	 * Ignore user provided values for period_length_ns and duty_offset_ns:
> +	 * we only support fixed period of MAX7360_PWM_PERIOD_NS and offset of

> +	 * 0.

Easy to read with 0 be on previous line.

> +	 */

> +

No need for this blank line.

> +	duty_steps = mul_u64_u64_div_u64(wf->duty_length_ns, MAX7360_PWM_MAX_RES,
> +					 MAX7360_PWM_PERIOD_NS);

This comes from math64.h

> +
> +	wfhw->duty_steps = min(MAX7360_PWM_MAX_RES, duty_steps);

...

> +static int max7360_pwm_write_waveform(struct pwm_chip *chip,
> +				      struct pwm_device *pwm,
> +				      const void *_wfhw)
> +{
> +	const struct max7360_pwm_waveform *wfhw = _wfhw;
> +	struct max7360_pwm *max7360_pwm;
> +	unsigned int val;
> +	int ret;
> +
> +	max7360_pwm = max7360_pwm_from_chip(chip);
> +
> +	val = (wfhw->duty_steps == 0) ? 0 : MAX7360_PWM_CTRL_ENABLE(pwm->hwpwm);
> +	ret = regmap_write_bits(max7360_pwm->regmap, MAX7360_REG_GPIOCTRL,
> +				MAX7360_PWM_CTRL_ENABLE(pwm->hwpwm), val);

> +
> +	if (!ret && wfhw->duty_steps != 0) {
> +		ret = regmap_write(max7360_pwm->regmap, MAX7360_REG_PWM(pwm->hwpwm),
> +				   wfhw->duty_steps);
> +	}
> +
> +	return ret;

Please, improve readability by rewriting like this:

	ret = regmap_write_bits(max7360_pwm->regmap, MAX7360_REG_GPIOCTRL,
				MAX7360_PWM_CTRL_ENABLE(pwm->hwpwm), val);
	if (ret)
		return ret;

	if (wfhw->duty_steps)
		return regmap_write(max7360_pwm->regmap, MAX7360_REG_PWM(pwm->hwpwm),
				    wfhw->duty_steps);

	return 0;

> +}

...

> +static int max7360_pwm_probe(struct platform_device *pdev)
> +{

With

	struct device *dev = &pdev->dev;

all below will look shorter and nicer.

> +	struct max7360_pwm *max7360_pwm;
> +	struct pwm_chip *chip;
> +	int ret;
> +
> +	if (!pdev->dev.parent)
> +		return dev_err_probe(&pdev->dev, -ENODEV, "no parent device\n");
> +
> +	chip = devm_pwmchip_alloc(pdev->dev.parent, MAX7360_NUM_PWMS,
> +				  sizeof(*max7360_pwm));
> +	if (IS_ERR(chip))
> +		return PTR_ERR(chip);
> +	chip->ops = &max7360_pwm_ops;
> +
> +	max7360_pwm = max7360_pwm_from_chip(chip);
> +	max7360_pwm->parent = pdev->dev.parent;
> +
> +	max7360_pwm->regmap = dev_get_regmap(pdev->dev.parent, NULL);
> +	if (!max7360_pwm->regmap)
> +		return dev_err_probe(&pdev->dev, -ENODEV,
> +				     "could not get parent regmap\n");

Will become one line (with the above suggestion).

> +	ret = devm_pwmchip_add(&pdev->dev, chip);

> +	if (ret != 0)

Please, be consistent with the style, and moreover this style is unusual.

> +		return dev_err_probe(&pdev->dev, ret,
> +				     "failed to add PWM chip\n");
> +
> +	return 0;
> +}

On Fri Feb 14, 2025 at 4:10 PM CET, Andy Shevchenko wrote:
> On Fri, Feb 14, 2025 at 12:49:53PM +0100, mathieu.dubois-briand@bootlin.com wrote:
> > From: Kamel Bouhara <kamel.bouhara@bootlin.com>
> > 
> > Add driver for Maxim Integrated MAX7360 PWM controller, supporting up to
> > 8 independent PWM outputs.
>
> ...
>
> > +#include <linux/of.h>
>
> Is this used? Cargo cult?
>

Right, not used anymore.

>
> > +struct max7360_pwm {
>
> > +	struct device *parent;
>
> Is it not the same as you can derive from regmap?
>

It is. I'm removing it.

Also, max7360_pwm structure will only contain the regmap, so I will
remove the structure and set the regmap directly as the pwm chip driver
data.

> ...
>

I have fixed all other points. Thanks for your review!

Hello Mathieu,

On Fri, Feb 14, 2025 at 12:49:53PM +0100, mathieu.dubois-briand@bootlin.com wrote:
> diff --git a/drivers/pwm/pwm-max7360.c b/drivers/pwm/pwm-max7360.c
> new file mode 100644
> index 000000000000..f1257c20add2
> --- /dev/null
> +++ b/drivers/pwm/pwm-max7360.c
> @@ -0,0 +1,213 @@
> +// SPDX-License-Identifier: GPL-2.0-only
> +/*
> + * Copyright 2024 Bootlin
> + *
> + * Author: Kamel BOUHARA <kamel.bouhara@bootlin.com>
> + * Author: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
> + *
> + * Limitations:
> + * - Only supports normal polarity.
> + * - The period is fixed to 2 ms.
> + * - Only the duty cycle can be changed, new values are applied at the beginning
> + *   of the next cycle.
> + * - When disabled, the output is put in Hi-Z.
> + */
> +#include <linux/err.h>
> +#include <linux/math.h>
> +#include <linux/mfd/max7360.h>
> +#include <linux/mod_devicetable.h>
> +#include <linux/module.h>
> +#include <linux/of.h>
> +#include <linux/platform_device.h>
> +#include <linux/pwm.h>
> +#include <linux/regmap.h>
> +
> +#define MAX7360_NUM_PWMS			8
> +#define MAX7360_PWM_MAX_RES			255
> +#define MAX7360_PWM_PERIOD_NS			2000000 /* 500 Hz */
> +#define MAX7360_PWM_COMMON_PWN			BIT(5)
> +#define MAX7360_PWM_CTRL_ENABLE(n)		BIT(n)
> +#define MAX7360_PWM_PORT(n)			BIT(n)
> +
> +struct max7360_pwm {
> +	struct device *parent;
> +	struct regmap *regmap;
> +};
> +
> +struct max7360_pwm_waveform {
> +	u8 duty_steps;
> +};
> +
> +static inline struct max7360_pwm *max7360_pwm_from_chip(struct pwm_chip *chip)
> +{
> +	return pwmchip_get_drvdata(chip);
> +}
> +
> +static int max7360_pwm_request(struct pwm_chip *chip, struct pwm_device *pwm)
> +{
> +	struct max7360_pwm *max7360_pwm;
> +	int ret;
> +
> +	max7360_pwm = max7360_pwm_from_chip(chip);
> +	ret = max7360_port_pin_request(max7360_pwm->parent, pwm->hwpwm, true);
> +	if (ret) {
> +		dev_warn(&chip->dev, "failed to request pwm-%d\n", pwm->hwpwm);

Please drop this warning, just returning ret here is fine. (The rule of
thumb is: Emit runtime messages only in probe, not during usage.)

> +		return ret;
> +	}
> +
> +	ret = regmap_write_bits(max7360_pwm->regmap,
> +				MAX7360_REG_PWMCFG(pwm->hwpwm),
> +				MAX7360_PWM_COMMON_PWN,
> +				0);
> +	if (ret)
> +		return ret;
> +
> +	return regmap_write_bits(max7360_pwm->regmap, MAX7360_REG_PORTS,
> +				 MAX7360_PWM_PORT(pwm->hwpwm),
> +				 MAX7360_PWM_PORT(pwm->hwpwm));
> +}
> +
> +static void max7360_pwm_free(struct pwm_chip *chip, struct pwm_device *pwm)
> +{
> +	struct max7360_pwm *max7360_pwm;
> +
> +	max7360_pwm = max7360_pwm_from_chip(chip);
> +	max7360_port_pin_request(max7360_pwm->parent, pwm->hwpwm, false);

Would be nice if pinmuxing would be abstracted as a pinctrl driver. Not
sure how much effort that is. Maybe Linus has an idea?

> +}
> +
> [...]
> +
> +static int max7360_pwm_write_waveform(struct pwm_chip *chip,
> +				      struct pwm_device *pwm,
> +				      const void *_wfhw)
> +{
> +	const struct max7360_pwm_waveform *wfhw = _wfhw;
> +	struct max7360_pwm *max7360_pwm;
> +	unsigned int val;
> +	int ret;
> +
> +	max7360_pwm = max7360_pwm_from_chip(chip);
> +
> +	val = (wfhw->duty_steps == 0) ? 0 : MAX7360_PWM_CTRL_ENABLE(pwm->hwpwm);

Does not setting MAX7360_PWM_CTRL_ENABLE result in the pin going to
Hi-Z? If yes: That's wrong. You're only supposed to do that if
period_length_ns = 0 was requested. If no: This needs a comment why
duty_steps = 0 is special here.

> +	ret = regmap_write_bits(max7360_pwm->regmap, MAX7360_REG_GPIOCTRL,
> +				MAX7360_PWM_CTRL_ENABLE(pwm->hwpwm), val);
> +
> +	if (!ret && wfhw->duty_steps != 0) {
> +		ret = regmap_write(max7360_pwm->regmap, MAX7360_REG_PWM(pwm->hwpwm),
> +				   wfhw->duty_steps);
> +	}
> +
> +	return ret;
> +}
> +
> +static int max7360_pwm_read_waveform(struct pwm_chip *chip,
> +				     struct pwm_device *pwm,
> +				     void *_wfhw)
> +{
> +	struct max7360_pwm_waveform *wfhw = _wfhw;
> +	struct max7360_pwm *max7360_pwm;
> +	unsigned int val;
> +	int ret;
> +
> +	max7360_pwm = max7360_pwm_from_chip(chip);
> +
> +	ret = regmap_read(max7360_pwm->regmap, MAX7360_REG_GPIOCTRL, &val);
> +	if (ret)
> +		return ret;
> +
> +	if (val & MAX7360_PWM_CTRL_ENABLE(pwm->hwpwm)) {
> +		ret = regmap_read(max7360_pwm->regmap, MAX7360_REG_PWM(pwm->hwpwm),
> +				  &val);
> +		val = wfhw->duty_steps;

wfhw->duty_steps = val;

> +	} else {
> +		wfhw->duty_steps = 0;
> +	}
> +
> +	return ret;
> +}

Best regards
Uwe