Received-SPF: pass (google.com: domain of linux-kernel+bounces-74604-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:40f1:3f00::1 as permitted sender) client-ip=2604:1380:40f1:3f00::1;
Date: Wed, 21 Feb 2024 11:20:07 +0000
From: Daniel Thompson <daniel.thompson@linaro.org>
To: Luca Weiss <luca@z3ntu.xyz>
Cc: ~postmarketos/upstreaming@lists.sr.ht, phone-devel@vger.kernel.org,
	Lee Jones <lee@kernel.org>, Jingoo Han <jingoohan1@gmail.com>,
	Helge Deller <deller@gmx.de>,
	Andrew Morton <akpm@linux-foundation.org>,
	"G.Shark Jeong" <gshark.jeong@gmail.com>,
	Bjorn Andersson <andersson@kernel.org>,
	Konrad Dybcio <konrad.dybcio@linaro.org>,
	Rob Herring <robh+dt@kernel.org>,
	Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
	Conor Dooley <conor+dt@kernel.org>,
	Maximilian Weigand <mweigand@mweigand.net>,
	dri-devel@lists.freedesktop.org, linux-fbdev@vger.kernel.org,
	linux-kernel@vger.kernel.org, linux-arm-msm@vger.kernel.org,
	devicetree@vger.kernel.org
Subject: Re: [PATCH 3/4] backlight: lm3630a: Use backlight_get_brightness
 helper in update_status
Message-ID: <20240221112007.GJ6716@aspen.lan>
References: <20240220-lm3630a-fixups-v1-0-9ca62f7e4a33@z3ntu.xyz>
 <20240220-lm3630a-fixups-v1-3-9ca62f7e4a33@z3ntu.xyz>
 <20240220141107.GF6716@aspen.lan>
 <5027630.31r3eYUQgx@g550jk>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <5027630.31r3eYUQgx@g550jk>

On Tue, Feb 20, 2024 at 05:43:32PM +0100, Luca Weiss wrote:
> On Dienstag, 20. Februar 2024 15:11:07 CET Daniel Thompson wrote:
> > On Tue, Feb 20, 2024 at 12:11:21AM +0100, Luca Weiss wrote:
> > > As per documentation "drivers are expected to use this function in their
> > > update_status() operation to get the brightness value.".
> > >
> > > With this we can also drop the manual backlight_is_blank() handling
> > > since backlight_get_brightness() is already handling this correctly.
> > >
> > > Signed-off-by: Luca Weiss <luca@z3ntu.xyz>
> >
> > Reviewed-by: Daniel Thompson <daniel.thompson@linaro.org>
> >
> > However...
> >
> > > ---
> > >  	/* disable sleep */
> > > @@ -201,9 +202,9 @@ static int lm3630a_bank_a_update_status(struct backlight_device *bl)
> > >  		goto out_i2c_err;
> > >  	usleep_range(1000, 2000);
> > >  	/* minimum brightness is 0x04 */
> > > -	ret = lm3630a_write(pchip, REG_BRT_A, bl->props.brightness);
> > > +	ret = lm3630a_write(pchip, REG_BRT_A, brightness);
> >
> > ... then handling of the minimum brightness looks weird in this driver.
> >
> > The range of the backlight is 0..max_brightness. Sadly the drivers
> > are inconsistant regarding whether zero means off or just minimum,
> > however three certainly isn't supposed to mean off! In other words the
> > offsetting should be handled by driver rather than hoping userspace has
> > some magic LM3630A mode.
>
> I could also try and fix that..
>
> 1. Treat 1..4 as 4, so have backlight on at that minimum level? Probably
> wouldn't be noticable that brightness 1=2=3=4. And the backlight will be
> on compared to off as it is now.
>
> 2. Decrease max_brightness by 4 values, so probably 0..251 and shift the
> values up in the driver so we get 4..255?
>
> Or would you have some other idea here?

I think #2 is the right option but shouldn't it be decrease max_brightness
by 3, yielding 0..252 .

Only nitpicking on that because, given how old this driver is I'd like
to be conservative. I don't expect there to be userspaces with magic
LM3630A modes but there could be some that assume #0 is off! Hence I
wanted to make sure we are on the same page.


Daniel.