Re: leds-pwm: issue in __led_pwm_set()

[Alexandre Belloni <alexandre.belloni@free-electrons.com>]

On 26/09/2013 02:28, Milo Kim wrote:
> Looping Thierry who maintains the PWM subsystem.
>
> I'm thinking your PWM driver is
> https://lkml.org/lkml/2013/9/24/776.
> Is it correct?
>

That is the one yes.

> However, I can't find any pin control for the PWM output.
> Just clock control and internal register access, PWM_EN and PWM_DIS.
> Is the pin controlled via the registers?
>

Pin control is taken care of by the pinctrl-at91 driver, shouldn't that
be enough ?

> Best regards,
> Milo
>
> On 09/26/2013 05:00 AM, Alexandre Belloni wrote:
>> Hi,
>>
>> I'm using leds-pwm on an atmel sama5d31 based board. I have one
>> question/issue. In __led_pwm_set(), if the duty cycle is 0,
>> pwm_disable() is called. This won't work fine on that platform. What
>> happens is that pwm_config() correctly sets the duty cycle to 0 which is
>> behaving correctly with that controller (and in my case, putting the
>> line low). But the call to pwm_disable() is making the pwm controller
>> release the line and then it is set to high.
>>
>> I've tried various configurations, like configuring a pull-down on the
>> pin but I still observe the same behavior. For now, I have a workaround
>> in atmel-pwm (I activate the output overdrive when duty cycle is 0). But
>> I believe, we should find another way to do that.
>>
>> Any input is appreciated.
>>
>> Regards,
>>
>
> -- 
> To unsubscribe from this list: send the line "unsubscribe linux-leds" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html


-- 
Alexandre Belloni, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com