Re: [PATCH 2/8] dt-bindings: pinctrl: aspeed: Convert AST2400 bindings to json-schema
From: Andrew Jeffery
Date: Thu Jun 27 2019 - 20:57:03 EST
On Fri, 28 Jun 2019, at 00:02, Rob Herring wrote:
> On Wed, Jun 26, 2019 at 9:55 PM Andrew Jeffery <andrew@xxxxxxxx> wrote:
> >
> >
> >
> > On Wed, 26 Jun 2019, at 23:17, Rob Herring wrote:
> > > On Wed, Jun 26, 2019 at 1:21 AM Andrew Jeffery <andrew@xxxxxxxx> wrote:
> > > > + The pin controller node should be the child of a syscon node with the
> > > > + required property:
> > > > +
> > > > + - compatible: Should be one of the following:
> > > > + "aspeed,ast2400-scu", "syscon", "simple-mfd"
> > > > + "aspeed,g4-scu", "syscon", "simple-mfd"
> > > > +
> > > > + Refer to the the bindings described in
> > > > + Documentation/devicetree/bindings/mfd/syscon.txt
> > > > +
> > > > + For the AST2400 pinmux, each mux function has only one associated pin group.
> > > > + Each group is named by its function. The following values for the function
> > > > + and groups properties are supported:
> > > > +
> > > > + ACPI ADC0 ADC1 ADC10 ADC11 ADC12 ADC13 ADC14 ADC15 ADC2 ADC3 ADC4 ADC5 ADC6
> > > > + ADC7 ADC8 ADC9 BMCINT DDCCLK DDCDAT EXTRST FLACK FLBUSY FLWP GPID GPID0 GPID2
> > > > + GPID4 GPID6 GPIE0 GPIE2 GPIE4 GPIE6 I2C10 I2C11 I2C12 I2C13 I2C14 I2C3 I2C4
> > > > + I2C5 I2C6 I2C7 I2C8 I2C9 LPCPD LPCPME LPCRST LPCSMI MAC1LINK MAC2LINK MDIO1
> > > > + MDIO2 NCTS1 NCTS2 NCTS3 NCTS4 NDCD1 NDCD2 NDCD3 NDCD4 NDSR1 NDSR2 NDSR3 NDSR4
> > > > + NDTR1 NDTR2 NDTR3 NDTR4 NDTS4 NRI1 NRI2 NRI3 NRI4 NRTS1 NRTS2 NRTS3 OSCCLK
> > > > + PWM0 PWM1 PWM2 PWM3 PWM4 PWM5 PWM6 PWM7 RGMII1 RGMII2 RMII1 RMII2 ROM16 ROM8
> > > > + ROMCS1 ROMCS2 ROMCS3 ROMCS4 RXD1 RXD2 RXD3 RXD4 SALT1 SALT2 SALT3 SALT4 SD1
> > > > + SD2 SGPMCK SGPMI SGPMLD SGPMO SGPSCK SGPSI0 SGPSI1 SGPSLD SIOONCTRL SIOPBI
> > > > + SIOPBO SIOPWREQ SIOPWRGD SIOS3 SIOS5 SIOSCI SPI1 SPI1DEBUG SPI1PASSTHRU
> > > > + SPICS1 TIMER3 TIMER4 TIMER5 TIMER6 TIMER7 TIMER8 TXD1 TXD2 TXD3 TXD4 UART6
> > > > + USB11D1 USB11H2 USB2D1 USB2H1 USBCKI VGABIOS_ROM VGAHS VGAVS VPI18 VPI24
> > > > + VPI30 VPO12 VPO24 WDTRST1 WDTRST2
> > >
> > > This should be a schema. You need to define child nodes and list these
> > > as values for 'function' and 'group'. Ideally, the child nodes would
> > > have some sort of pattern, but if not, you can just match on '^.*$'
> > > under patternProperties.
> >
> > The children don't have any pattern in their node name, which drives
> > me towards the '^.*$' pattern match, however, what I've found is that
> > I get the following errors for some of the relevant dts files:
> >
> > ```
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-opp-palmetto.dt.yaml: compatible: ['aspeed,g4-pinctrl'] is not of type 'object'
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-opp-palmetto.dt.yaml: pinctrl-names: ['default'] is not of type 'object'
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-opp-palmetto.dt.yaml: pinctrl-0: [[7, 8, 9, 10, 11, 12]] is not of type 'object'
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-opp-palmetto.dt.yaml: phandle: [[13]] is not of type 'object'
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-opp-palmetto.dt.yaml: $nodename: ['pinctrl'] is not of type 'object'
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-quanta-q71l.dt.yaml: compatible: ['aspeed,g4-pinctrl'] is not of type 'object'
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-quanta-q71l.dt.yaml: pinctrl-names: ['default'] is not of type 'object'
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-quanta-q71l.dt.yaml: pinctrl-0: [[9, 10, 11, 12]] is not of type 'object'
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-quanta-q71l.dt.yaml: phandle: [[13]] is not of type 'object'
> > /home/andrew/src/linux/aspeed/arch/arm/boot/dts/aspeed-bmc-quanta-q71l.dt.yaml: $nodename: ['pinctrl'] is not of type 'object'
> > ```
> >
>
> The problem is "^.*$" matches both properties and child nodes.
>
> > We shouldn't be expecting these properties in the child nodes, so
> > something is busted. Looking at processed-schema.yaml, we have:
> >
> > ```
> > - $filename: /home/andrew/src/linux/aspeed/Documentation/devicetree/bindings/pinctrl/aspeed,ast2400-pinctrl.yaml
> > $id: http://devicetree.org/schemas/pinctrl/aspeed,ast2400-pinctrl.yaml#
> > $schema: http://devicetree.org/meta-schemas/core.yaml#
> > patternProperties:
> > ^.*$:
> > patternProperties:
> > ^function|groups$:
> > allOf:
> > - {$ref: /schemas/types.yaml#/definitions/string}
> > - additionalItems: false
> > items:
> > enum: [ACPI, ADC0, ADC1, ADC10, ADC11, ADC12, ADC13, ADC14, ADC15, ADC2,
> > ADC3, ADC4, ADC5, ADC6, ADC7, ADC8, ADC9, BMCINT, DDCCLK, DDCDAT,
> > EXTRST, FLACK, FLBUSY, FLWP, GPID, GPID0, GPID2, GPID4, GPID6, GPIE0,
> > GPIE2, GPIE4, GPIE6, I2C10, I2C11, I2C12, I2C13, I2C14, I2C3, I2C4,
> > I2C5, I2C6, I2C7, I2C8, I2C9, LPCPD, LPCPME, LPCRST, LPCSMI, MAC1LINK,
> > MAC2LINK, MDIO1, MDIO2, NCTS1, NCTS2, NCTS3, NCTS4, NDCD1, NDCD2,
> > NDCD3, NDCD4, NDSR1, NDSR2, NDSR3, NDSR4, NDTR1, NDTR2, NDTR3, NDTR4,
> > NDTS4, NRI1, NRI2, NRI3, NRI4, NRTS1, NRTS2, NRTS3, OSCCLK, PWM0,
> > PWM1, PWM2, PWM3, PWM4, PWM5, PWM6, PWM7, RGMII1, RGMII2, RMII1, RMII2,
> > ROM16, ROM8, ROMCS1, ROMCS2, ROMCS3, ROMCS4, RXD1, RXD2, RXD3, RXD4,
> > SALT1, SALT2, SALT3, SALT4, SD1, SD2, SGPMCK, SGPMI, SGPMLD, SGPMO,
> > SGPSCK, SGPSI0, SGPSI1, SGPSLD, SIOONCTRL, SIOPBI, SIOPBO, SIOPWREQ,
> > SIOPWRGD, SIOS3, SIOS5, SIOSCI, SPI1, SPI1DEBUG, SPI1PASSTHRU, SPICS1,
> > TIMER3, TIMER4, TIMER5, TIMER6, TIMER7, TIMER8, TXD1, TXD2, TXD3,
> > TXD4, UART6, USB11D1, USB11H2, USB2D1, USB2H1, USBCKI, VGABIOS_ROM,
> > VGAHS, VGAVS, VPI18, VPI24, VPI30, VPO12, VPO24, WDTRST1, WDTRST2]
> > maxItems: 1
> > minItems: 1
> > type: array
> > pinctrl-[0-9]+: true
> > properties: {phandle: true, pinctrl-names: true, status: true}
> > type: object
> > pinctrl-[0-9]+: true
> > properties:
> > $nodename: true
> > compatible:
> > additionalItems: false
> > items:
> > - enum: ['aspeed,ast2400-pinctrl', 'aspeed,g4-pinctrl']
> > maxItems: 1
> > minItems: 1
> > type: array
> > phandle: true
> > pinctrl-names: true
> > status: true
> > required: [compatible]
> > select:
> > properties:
> > compatible:
> > contains:
> > enum: ['aspeed,ast2400-pinctrl', 'aspeed,g4-pinctrl']
> > required: [compatible]
> > title: ASPEED AST2400 Pin Controller
> > ```
> >
> > `properties: {phandle: true, pinctrl-names: true, status: true}` has been
> > merged into my '^.*$' patternProperty, presumably partly from
> > pinctrl-consumer.yaml, and this seems to be the source of the bad
> > output. If as a hack I change my pattern to '^.*_default$' the problem
> > goes away as we no longer try to enforce the constraints on properties
> > provided by other bindings, but the problem is the node names are
> > largely freeform[1] (unless I enforce a naming constraint as part of my
> > bindings?).
> >
> > [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/devicetree/bindings/pinctrl/pinctrl-bindings.txt?h=v5.2-rc6#n112
> >
> > >
> > > BTW, You can put the names under a 'definitions' key and then use
> > > '$ref' to reference them from function and group to avoid duplicating
> > > the names. Or use patternProperties with '^(function|group)$'.
> >
> > I've used the patternProperties approach above as I couldn't get the
> > definitions/$ref approach to work. I did the following:
>
> The problem is we'd need to process the schema under definitions. The
> YAML encoding we validate against always encodes strings as arrays as
> dtc has no way of knowing if a given property is a string array or
> single string. So to avoid a bunch of boilerplate in every binding, we
> process the schema to transform single strings into arrays of length
> 1.
>
> It's probably best to stick with the patternProperties approach. I
> think you can do something like this:
>
> "^.*$":
> if:
> type: object
> then:
> patternProperties:
> '^(function|group)$':
> ...
>
> I'm not completely certain this works though, so if you can send me an
> updated binding with what you have so far I can test it out.
This works, thanks. I'll send an updated series.
Andrew