Re: [PATCH v8 1/2] PWM: atmel-pwm: add PWM controller driver

From: Thierry Reding
Date: Fri Dec 06 2013 - 08:31:27 EST


On Thu, Dec 05, 2013 at 09:11:28AM +0800, Bo Shen wrote:
> Hi Thierry,
>
> On 12/04/2013 06:03 PM, Thierry Reding wrote:
> >On Wed, Dec 04, 2013 at 10:59:46AM +0800, Bo Shen wrote:
> >>Hi Thierry,
> >>
> >>On 12/03/2013 05:43 PM, Thierry Reding wrote:
> >>>On Tue, Dec 03, 2013 at 11:09:12AM +0800, Bo Shen wrote:
> >>>>On 12/02/2013 06:59 PM, Thierry Reding wrote:
> >>>>>On Mon, Nov 18, 2013 at 05:13:21PM +0800, Bo Shen wrote:
> >>>[...]
> >>>>>>diff --git a/drivers/pwm/pwm-atmel.c b/drivers/pwm/pwm-atmel.c
> >>>[...]
> >>>>>>+ /* Calculate the period cycles */
> >>>>>>+ while (div > PWM_MAX_PRD) {
> >>>>>>+ div = clk_rate / (1 << pres);
> >>>>>>+ div = div * period_ns;
> >>>>>>+ /* 1/Hz = 100000000 ns */
> >>>>>
> >>>>>I don't think that comment is needed.
> >>>>
> >>>>This is asked to be added.
> >>>>And, I think keep it and it won't hurt, what do you think?
> >>>
> >>>I think it's obvious that you're converting from nanoseconds because of
> >>>the _ns prefix in period_ns. But if somebody requested this and everyone
> >>>else thinks it's useful, I'm okay with keeping it.
> >>>
> >>>>>>+ if (test_bit(PWMF_ENABLED, &pwm->flags)) {
> >>>>>>+ atmel_pwm_ch_writel(atmel_pwm, pwm->hwpwm, PWMV2_CDTYUPD, dty);
> >>>>>>+ } else {
> >>>>>>+ atmel_pwm_ch_writel(atmel_pwm, pwm->hwpwm, PWMV2_CDTY, dty);
> >>>>>>+ atmel_pwm_ch_writel(atmel_pwm, pwm->hwpwm, PWMV2_CPRD, prd);
> >>>>>>+ }
> >>>>>>+}
> >>>>>
> >>>>>Neither version 1 nor version 2 seem to be able to change the period
> >>>>>while the channel is enabled. Perhaps that should be checked for in
> >>>>>atmel_pwm_config() and an error (-EBUSY) returned?
> >>>>
> >>>>The period is configured in dt in device tree, or platform data in non
> >>>>device tree. Nowhere will update period. So, not code to update period.
> >>>>Am I right? If not, please figure out.
> >>>
> >>>The .config() operation allows the period to be specified. Just because
> >>>nobody currently changes it at runtime doesn't mean it can't be done.
> >>>
> >>>It is also possible that whoever wrote the device tree or platform data
> >>>didn't know that the period must be the same for all PWM channels and
> >>>therefore might use different values. If you check for this, at least
> >>>they'll notice. If you don't check for it, then they may end up having
> >>>the period reconfigured behind their backs, which may cause parts of
> >>>their setup to behave unexpectedly.
> >>
> >>Thanks for this information.
> >>I will add code for changing period.
> >
> >Just to clarify: I wouldn't want this code to allow changing the period
> >but rather reject incompatible changes to the period with an error code.
>
> So, in this patch, just check it as you suggested in previous email, would
> it be OK?
> --->8---
> Perhaps that should be checked for in atmel_pwm_config() and an error
> (-EBUSY) returned?
> ---8<---

Yes. If a user tries to set a period that conflicts with a previously
set period (by another PWM channel), then pwm_config() for the second
user should fail.

Thierry

Attachment: pgp00000.pgp
Description: PGP signature