linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>
To: Miquel Raynal <miquel.raynal@bootlin.com>
Cc: Thierry Reding <thierry.reding@gmail.com>,
	Linus Walleij <linus.walleij@linaro.org>,
	Bartosz Golaszewski <bgolaszewski@baylibre.com>,
	linux-gpio@vger.kernel.org, linux-pwm@vger.kernel.org,
	Andy Shevchenko <andy.shevchenko@gmail.com>,
	Thomas Petazzoni <thomas.petazzoni@bootlin.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v6] gpio: pca953x: Add Maxim MAX7313 PWM support
Date: Fri, 3 Jul 2020 16:53:13 +0200	[thread overview]
Message-ID: <20200703145313.vwjsh5crdqx2u76a@pengutronix.de> (raw)
In-Reply-To: <20200503105453.23658-1-miquel.raynal@bootlin.com>

[-- Attachment #1: Type: text/plain, Size: 4201 bytes --]

Hello Miquel,

On Sun, May 03, 2020 at 12:54:53PM +0200, Miquel Raynal wrote:
> +static u8 max7313_pwm_get_intensity(struct pca953x_chip *pca_chip,
> +				    unsigned int idx)
> +{
> +	struct device *dev = &pca_chip->client->dev;
> +	unsigned int reg, shift, val, output;
> +	u8 intensity;
> +	bool phase;
> +	int ret;
> +
> +	/* Retrieve the intensity */
> +	reg = MAX7313_INTENSITY + (idx / PWM_PER_REG);
> +	shift = (idx % PWM_PER_REG) ? PWM_BITS_PER_REG : 0;

I would find

	shift = (idx % PWM_PER_REG) * PWM_BITS_PER_REG

more natural here as your formula only works for PWM_PER_REG = 2.

> +	mutex_lock(&pca_chip->i2c_lock);
> +	ret = regmap_read(pca_chip->regmap, reg, &val);
> +	mutex_unlock(&pca_chip->i2c_lock);
> +	if (ret < 0) {
> +		dev_err(dev, "Cannot retrieve PWM intensity (%d)\n", ret);

Please use %pe for error codes.

> +		return 0;
> +	}
> +
> +	val >>= shift;
> +	val &= PWM_INTENSITY_MASK;
> +
> +	/* Retrieve the phase */
> +	reg = pca953x_recalc_addr(pca_chip, pca_chip->regs->output, idx, 0, 0);
> +
> +	mutex_lock(&pca_chip->i2c_lock);
> +	ret = regmap_read(pca_chip->regmap, reg, &output);
> +	mutex_unlock(&pca_chip->i2c_lock);
> +	if (ret < 0) {
> +		dev_err(dev, "Cannot retrieve PWM phase (%d)\n", ret);
> +		return 0;
> +	}
> +
> +	phase = output & BIT(idx % BANK_SZ);

Would it make sense to cache the phase value to reduce register access
and locking here?

> [...]
> +static int max7313_pwm_apply(struct pwm_chip *chip,
> +			     struct pwm_device *pwm,
> +			     const struct pwm_state *state)
> +{
> +	struct max7313_pwm *max_pwm = to_max7313_pwm(chip);
> +	struct pca953x_chip *pca_chip = to_pca953x(max_pwm);
> +	unsigned int intensity, active;
> +	int ret = 0;
> +
> +	if (!state->enabled ||
> +	    state->period < PWM_PERIOD_NS ||
> +	    state->polarity != PWM_POLARITY_NORMAL)
> +		return -EINVAL;

You could simulate state->enabled = false using duty_cycle = 0.

> +	/* Convert the duty-cycle to be in the [0;16] range */
> +	intensity = max7313_pwm_duty_to_intensity(state->duty_cycle);

This might return a value > 16 if state->duty_cycle > PWM_PERIOD_NS.
I suggest to do

	duty_cycle = min(state->duty_cycle, PWM_PERIOD_NS);

and use that value instead of state->duty_cycle.

> +	/*
> +	 * The hardware is supposedly glitch-free when changing the intensity,
> +	 * unless we need to flip the blink phase to reach an extremity or the
> +	 * other of the spectrum (0/16 from phase 1, 16/16 from phase 0).

s/other of/other end of/. I don't understand the difference between
extremity and "other end of the spectrum".

> +	 */
> +	return max7313_pwm_set_state(pca_chip, pwm, intensity);
> +}
> +
> +static void max7313_pwm_get_state(struct pwm_chip *chip,
> +				  struct pwm_device *pwm,
> +				  struct pwm_state *state)
> +{
> +	struct max7313_pwm *max_pwm = to_max7313_pwm(chip);
> +	struct pca953x_chip *pca_chip = to_pca953x(max_pwm);
> +	u8 intensity;
> +
> +	state->enabled = true;
> +	state->period = PWM_PERIOD_NS;
> +	state->polarity = PWM_POLARITY_NORMAL;
> +
> +	intensity = max7313_pwm_get_intensity(pca_chip, pwm->hwpwm);
> +	state->duty_cycle = max7313_pwm_intensity_to_duty(intensity);

Please round up the division in max7313_pwm_intensity_to_duty().

(The reasoning is: with rounding down the following can happen:

	/* this configures for 15/16 */
	pwm_apply_state(pwm, { .duty_cycle = 31249, .period = 31250 });

	/* assume this called your .get_state callback */
	pwm_get_state(pwm, &state);

	/*
	 * we now have
	 *	state.duty_cycle = 29296;
	 *	state.period = 31250;
	 * right?
	 */
	
	/* this configures for 14/16 because 29296 * 16 / 31250 < 15 */
	pwm_apply_state(pwm, &state);

But I want this to be idempotent, i.e. pwm_get_state has to round up and
then return .duty_cycle = 29297 in the above example which is enough to
let .apply_state() configure 15/16 again. Enabling PWM_DEBUG should
catch this.)

> +};

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | https://www.pengutronix.de/ |

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

  parent reply	other threads:[~2020-07-03 14:53 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-03 10:54 [PATCH v6] gpio: pca953x: Add Maxim MAX7313 PWM support Miquel Raynal
2020-05-03 19:20 ` Andy Shevchenko
2020-05-03 19:41   ` Uwe Kleine-König
2020-06-29 14:08 ` Miquel Raynal
2020-06-29 16:50   ` Andy Shevchenko
2020-06-29 19:50   ` Uwe Kleine-König
2020-06-30  6:58     ` Miquel Raynal
2020-06-30  9:08       ` Uwe Kleine-König
2020-06-30  9:13       ` Bartosz Golaszewski
2020-06-30 12:45         ` Uwe Kleine-König
2020-06-30 21:27           ` Andy Shevchenko
2020-06-30 21:56             ` Miquel Raynal
2020-07-03 14:53 ` Uwe Kleine-König [this message]
2020-07-04 11:23   ` Andy Shevchenko
2020-07-04 15:43     ` Uwe Kleine-König
2020-08-12 17:36   ` Miquel Raynal

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200703145313.vwjsh5crdqx2u76a@pengutronix.de \
    --to=u.kleine-koenig@pengutronix.de \
    --cc=andy.shevchenko@gmail.com \
    --cc=bgolaszewski@baylibre.com \
    --cc=linus.walleij@linaro.org \
    --cc=linux-gpio@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pwm@vger.kernel.org \
    --cc=miquel.raynal@bootlin.com \
    --cc=thierry.reding@gmail.com \
    --cc=thomas.petazzoni@bootlin.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).