linux-leds.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Milo Kim <milo.kim@ti.com>
To: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Cc: linux-leds@vger.kernel.org,
	Thierry Reding <thierry.reding@gmail.com>,
	Nicolas Ferre <nicolas.ferre@atmel.com>,
	Bo Shen <voice.shen@atmel.com>
Subject: Re: leds-pwm: issue in  __led_pwm_set()
Date: Sat, 28 Sep 2013 02:23:14 +0900	[thread overview]
Message-ID: <5245BF02.80804@ti.com> (raw)
In-Reply-To: <5245AF24.9070700@free-electrons.com>

On 09/28/2013 01:15 AM, Alexandre Belloni wrote:
> Cc the atmel maintainers that may want to chime in.
>
> On 27/09/2013 17:51, Milo Kim wrote:
>> On 09/26/2013 04:48 PM, Alexandre Belloni wrote:
>>>
>>> Hum, maybe my wording was wrong. What I meant is that when that
>>> disabling the PWM channel by using the PWM_DIS
>>>    register, the PWM is not driving the pin anymore. Then, the level goes
>>> from low (which is correct) to high. Also, the datasheet specifies that
>>> the pwm has to be enabled to get the correct level when duty == 0 or
>>> duty == period.
>>>
>>> IIRC, this is not the same on TI SoC where duty == 0 don't give you the
>>> expected behavior and I can understand why there is a pwm_disable()
>>> there. Maybe we have to have a way to differentiate both cases ?
>>>
>> Based on your result, PWM_DIS should be updated when the driver is
>> unloaded - no PWM consumer anymore.
>>
>> Why don't you move PWM_DIS register access code from
>> atmel_pwm_disable() to atmel_pwm_remove()?
>> If it makes sense, the PWM_EN code also needs to be moved to _probe().
>>
>
> If we do what you suggest, I'm afraid we will enable pwm channels that
> have no consumers. Isn't pwm_disable() suppose to disable the pwm
> channels ? I believe that is what is done on the other platforms.
>

Ah, I missed that point, thanks.
What about using pwm_request() and pwm_free()?

1) Add pwm_request() and pwm_free ops in the atmel-pwm driver.

static const struct pwm_ops atmel_pwm_ops = {
+	.request = atmel_pwm_request,
+	.free = atmel_pwm_free,
	.config = atmel_pwm_config,
	.set_polarity = atmel_pwm_set_polarity,
	.enable = atmel_pwm_enable,
	.disable = atmel_pwm_disable,
	.owner = THIS_MODULE,
};

2) Move atmel PWM register code to atmel_pwm_request() and _free()

static int atmel_pwm_request(struct pwm_chip *chip, struct pwm_device *pwm)
{
	struct atmel_pwm_chip *atmel_pwm = to_atmel_pwm_chip(chip);

	atmel_pwm_writel(atmel_pwm, PWM_ENA, 1 << pwm->hwpwm);
	return 0;
}

static void atmel_pwm_disable(struct pwm_chip *chip, struct pwm_device *pwm)
{
	struct atmel_pwm_chip *atmel_pwm = to_atmel_pwm_chip(chip);

	atmel_pwm_writel(atmel_pwm, PWM_DIS, 1 << pwm->hwpwm);
}

It still has some issue which you pointed if a PWM consumer doesn't call 
pwm_enable() after pwm_get().
However, it's better idea than enabling the PWM channel at initial time.

Best regards,
Milo

  reply	other threads:[~2013-09-27 17:23 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-09-25 20:00 leds-pwm: issue in __led_pwm_set() Alexandre Belloni
2013-09-26  0:28 ` Milo Kim
2013-09-26  7:13   ` Alexandre Belloni
2013-09-26  7:34     ` Milo Kim
2013-09-26  7:48       ` Alexandre Belloni
2013-09-27 15:51         ` Milo Kim
2013-09-27 16:15           ` Alexandre Belloni
2013-09-27 17:23             ` Milo Kim [this message]
2013-09-27 17:33               ` Milo Kim
2013-09-27 18:34                 ` Alexandre Belloni
2013-09-27 19:46                   ` Alexandre Belloni
2013-09-28 13:29                     ` Milo Kim

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=5245BF02.80804@ti.com \
    --to=milo.kim@ti.com \
    --cc=alexandre.belloni@free-electrons.com \
    --cc=linux-leds@vger.kernel.org \
    --cc=nicolas.ferre@atmel.com \
    --cc=thierry.reding@gmail.com \
    --cc=voice.shen@atmel.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).