Re: [PATCH 3/3] arm64: dts: qcom: pm8350c: Add pwm support

From: skakit@codeaurora.org
To: Bjorn Andersson <bjorn.andersson@linaro.org>,
	Matthias Kaehlcke <mka@chromium.org>
Cc: Pavel Machek <pavel@ucw.cz>, Rob Herring <robh+dt@kernel.org>,
	Andy Gross <agross@kernel.org>,
	swboyd@chromium.org, kgunda@codeaurora.org,
	linux-leds@vger.kernel.org, devicetree@vger.kernel.org,
	linux-kernel@vger.kernel.org, linux-arm-msm@vger.kernel.org
Subject: Re: [PATCH 3/3] arm64: dts: qcom: pm8350c: Add pwm support
Date: Thu, 09 Sep 2021 11:41:29 +0530	[thread overview]
Message-ID: <667da76fe38a67af196e8dad00decd71@codeaurora.org> (raw)
In-Reply-To: <YTjuGVOmzkTot8z1@ripper>

On 2021-09-08 22:38, Bjorn Andersson wrote:
> On Wed 08 Sep 08:29 PDT 2021, Matthias Kaehlcke wrote:
> 
>> On Wed, Sep 08, 2021 at 02:37:39PM +0530, skakit@codeaurora.org wrote:
>> > On 2021-09-07 23:46, Matthias Kaehlcke wrote:
>> > > On Mon, Sep 06, 2021 at 04:11:07PM +0530, satya priya wrote:
>> > > > Add pwm support for PM8350C pmic.
>> > > >
>> > > > Signed-off-by: satya priya <skakit@codeaurora.org>
>> > > > ---
>> > > >  arch/arm64/boot/dts/qcom/pm8350c.dtsi | 6 ++++++
>> > > >  1 file changed, 6 insertions(+)
>> > > >
>> > > > diff --git a/arch/arm64/boot/dts/qcom/pm8350c.dtsi
>> > > > b/arch/arm64/boot/dts/qcom/pm8350c.dtsi
>> > > > index e1b75ae..ecdae55 100644
>> > > > --- a/arch/arm64/boot/dts/qcom/pm8350c.dtsi
>> > > > +++ b/arch/arm64/boot/dts/qcom/pm8350c.dtsi
>> > > > @@ -29,6 +29,12 @@
>> > > >  			interrupt-controller;
>> > > >  			#interrupt-cells = <2>;
>> > > >  		};
>> > > > +
>> > > > +		pm8350c_pwm4: pwm {
>> > >
>> > > What does the '4' represent, an internal channel number? It should
>> > > probably be omitted if the PM8350 only has a single output PWM
>> > > port.
>> > >
>> >
>> > pm8350c has four PWMs, but I think we can drop the '4' here.
>> 
>> Why is only one PWM exposed if the PMIC has for of them? Why number 4
>> and not one of the others?
> 

pwm4 is used for backlight support on kodiak crd board, so I mentioned 
4, thinking 4 nodes should be present for 4 pwms.
but I see that we need to represent all the four channels as one node. 
will drop the '4' in next version.

Thanks,
Satya Priya

> The node should represent all 4 channels, which ones the board uses is
> captured in how they are bound to other clients - or defines as LEDs by
> additional child nodes.
> 
> Regards,
> Bjorn