From mboxrd@z Thu Jan 1 00:00:00 1970 From: Lee Jones Subject: Re: [PATCH V4 1/5] DT: mfd: add device-tree binding doc fro PMIC max77620/max20024 Date: Tue, 26 Jan 2016 14:58:54 +0000 Message-ID: <20160126145854.GC3368@x1> References: <1453198783-28383-1-git-send-email-ldewangan@nvidia.com> <1453198783-28383-2-git-send-email-ldewangan@nvidia.com> <20160125115610.GC3368@x1> <56A63A85.6060609@nvidia.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Content-Disposition: inline In-Reply-To: <56A63A85.6060609@nvidia.com> Sender: linux-gpio-owner@vger.kernel.org To: Laxman Dewangan Cc: robh+dt@kernel.org, pawel.moll@arm.com, mark.rutland@arm.com, ijc+devicetree@hellion.org.uk, galak@codeaurora.org, linus.walleij@linaro.org, gnurou@gmail.com, broonie@kernel.org, a.zummo@towertech.it, alexandre.belloni@free-electrons.com, lgirdwood@gmail.com, devicetree@vger.kernel.org, linux-kernel@vger.kernel.org, linux-gpio@vger.kernel.org, rtc-linux@googlegroups.com, swarren@nvidia.com, treding@nvidia.com, k.kozlowski@samsung.com, vreddytalla@nvidia.com List-Id: devicetree@vger.kernel.org On Mon, 25 Jan 2016, Laxman Dewangan wrote: >=20 > Thanks Lee for your review. I will take care of most of comment on > my next patch. >=20 > I have reply on some comment and seeking more details for few > comments as follows. >=20 > On Monday 25 January 2016 05:26 PM, Lee Jones wrote: > >On Tue, 19 Jan 2016, Laxman Dewangan wrote: > > > >>+- interrupt-controller: MAX77620 has internal interrupt controller= which > >>+ takes the interrupt request from internal sub-blocks like RTC, > >>+ regulators, GPIOs as well as external input. > >This is how interrupt-controllers usually work. I don't think there > >is any need to explain this. I'd just link to > >../interrupt-controller/interrupts.txt instead. >=20 > Something similar to (just to confirm) > - interrupt-controller: describes the 88pm860x as an interrupt > controller (has its own domain) IRQ domains are Linuxisums, which aren't allowed in DT. I'd prefer something along the lines of: "Identifies the node as an interrupt controller." > >>+- #interrupt-cells: Should be set to 2 for IRQ number and flags. > >>+ The first cell is the IRQ number. IRQ numbers for different inte= rrupt > >>+ source of MAX77620 are defined at dt-bindings/mfd/max77620.h > >>+ The second cell is the flags, encoded as the trigger masks from = binding > >>+ document interrupts.txt, using dt-bindings/irq. > >This is a very lengthy read for such little information. Please mak= e > >it more succinct. Take a look at other files for examples. > I started with as3722.txt and it seems I am carrying forward the > complexity here. Originally, as3722 is posted by me only. Any good > file example which I can refer? Just grep for #interrupt-cells in Documentation. There are plenty of nice, short examples. > > Also, > >tell use where interrupts.txt is > >i.e. ../interrupt-controller/interrupts.txt. > > > >These are so much easier to read if you tab out from the property na= me > >to the description. > > > >- reg: I2C device address. > >- interrupt-controller: MAX77620 has internal interrupt controller = which > > takes the interrupt request from internal > > sub-blocks like RTC, regulators, GPIOs as well > > as external input. > >- #interrupt-cells: Should be set to 2 for IRQ number and flags. > > The first cell is the IRQ number. IRQ numbers > > for different interrupt source of MAX77620 are > > defined at dt-bindings/mfd/max77620.h > > The second cell is the flags, encoded as the > > trigger masks from binding document > > interrupts.txt, using dt-bindings/irq. > > > >... don't you think? >=20 > Agree, I can make indenting. And will do whatever places it needs in > this document. >=20 > > > >>+Optional properties: > >>+------------------- > >>+This device also supports the power OFF of system. > >What is the "power OFF of system"? >=20 > PMIC supplies the power. So once PMIC is in OFF state, it turns off > all its rail and hence no SW execution on system. Still external > supply will keep supply to PMIC. "This device controls platform power" or something. > >>+Following properties are used for this purpose: > >>+- system-power-controller: Boolean, This device will be use as > >You don't describe the type of each property, so why is this one > >special? > Hmm. I describe the boolean and tristate only. Do I need to define > type for integer,string also? I wouldn't describe any of them. It's normally pretty obvious which properties are boolean by the lack of required cell description. > >>+power OFF slot and slot period of the device. Device has 3 FPS as = =46PS0, > >>+FPS1 and FPS2. The details of FPS configuration is provided throug= h > >>+subnode "fps". The details of FPS0, FPS1, FPS2 are provided throug= h the > >>+child node under this subnodes. The FPS number is provided via reg= property. > >>+ > >>+The property for fps child nodes as: > >>+Required properties: > >>+ -reg: FPS number like 0, 1, 2 for FPS0, FPS1 and FPS2 respectivel= y. > >I'm surprised Rob Acked this. We don't usually do device numbers in= DT. > What is best way to make the child node for FPS and differentiate FPS= 0,1, 2? > What is your suggestion here? There are lots of ways you can solve this and so many examples of others doing so. I suggest you have a look at some DTS files and figure it out. One possible solution is to use different compatible strings. > >>+ -maxim,active-fps-time-period-us: Active state FPS time period in > >>+ microseconds. > >>+ -maxim,suspend-fps-time-period-us: Suspend state FPS time period = in > >>+ microseconds. > >>+ -maxim,fps-enable-input: FPS enable source like EN0, EN1 or SW. T= he > >>+ macros are defined on dt-bindings/mfd/max77620.h for > >>+ different enable source. > >>+ FPS_EN_SRC_EN0 for EN0 enable source. > >>+ FPS_EN_SRC_EN1 for En1 enable source. > >>+ FPS_EN_SRC_SW for SW based control. > >>+ -maxim,fps-sw-enable: Boolean, applicable if enable input is SW. > >>+ If this property present then enable the FPS else > >"property is present" > > > >If this enables/disables FPS, why does it matter if it's SW or not? > >Why can't you just cal it maxim,fps-enable? Also, is there a case > >where you would supply this sub-node, have FPS enabled and this > >property not present? If not, can't you just remove the entire node= ? > >Or am I missing something? >=20 > Here, my need is to connect different FPS source inputs EN0, EN1 or > SW. They can connected to any inputs. That's why fps-enable-input > select the related digital input for given FPS. > However, I can reduce the need of "fps-sw-enable" and make this > always enable if fps-enable-input=3D . This is exactly how you need to think. Reducing the number of properties you introduce (and this device adds a lot more than I would normally expect by the way) needs to be top priority. > Here is excerpt from datasheet: > B3:B1: SRCFPSx[1:0] > Enable Source. Specifies the enable source for the sequencer. > 0b00=3DEN0 hardware input > 0b01=3DEN1 hardware input > 0b10=3DENFPSx software bit > 0b11=3Dreserved > B0 ENFPSx > Software Enable > 0=3DDisable FPSx > 1=3DEnable FPSx > X=3DENFPSx is a don=C2=92t care if SRCFPSx[1:0] !=3D 0b10 >=20 >=20 > > > >>+ -maxim,enable-global-lpm: Boolean, enable global LPM when the ext= ernal > >LPM? > Low Power Mode (LPM). I will add details. >=20 > >+Pinmux and GPIO: > >+=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > >I think this whole section needs moving to ../pinctrl and needs to b= e > >reviewed by Linus W. >=20 > Is this mean I need to create DT binding doc for the each subsystem > differently? > Actually during AS3722, I had different understanding to have single = file. Yes, that way you have each of the the subsystem experts review your documentation. You can then link to them from this document. --=20 Lee Jones Linaro STMicroelectronics Landing Team Lead Linaro.org =E2=94=82 Open source software for ARM SoCs =46ollow Linaro: Facebook | Twitter | Blog -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html