[PATCH 05/10] mips: dts: ralink: mt7621: fix some dtc warnings
Arınç ÜNAL
arinc.unal at arinc9.com
Wed Sep 14 23:24:39 PDT 2022
On 15.09.2022 08:59, Sergio Paracuellos wrote:
> On Thu, Sep 15, 2022 at 5:30 AM Sergio Paracuellos
> <sergio.paracuellos at gmail.com> wrote:
>>
>> On Wed, Sep 14, 2022 at 12:46 PM Arınç ÜNAL <arinc.unal at arinc9.com> wrote:
>>>
>>> Hi Sergio,
>>>
>>> On 14.09.2022 12:14, Sergio Paracuellos wrote:
>>>> Hi Arinc,
>>>>
>>>> On Wed, Sep 14, 2022 at 10:55 AM Arınç ÜNAL <arinc.unal at arinc9.com> wrote:
>>>>>
>>>>> Fix the dtc warnings below.
>>>>>
>>>>> /cpus/cpu at 0: failed to match any schema with compatible: ['mips,mips1004Kc']
>>>>> /cpus/cpu at 1: failed to match any schema with compatible: ['mips,mips1004Kc']
>>>>> uartlite at c00: $nodename:0: 'uartlite at c00' does not match '^serial(@.*)?$'
>>>>> From schema: /home/arinc9/Documents/linux/Documentation/devicetree/bindings/serial/8250.yaml
>>>>> uartlite at c00: Unevaluated properties are not allowed ('clock-names' was unexpected)
>>>>> From schema: /home/arinc9/Documents/linux/Documentation/devicetree/bindings/serial/8250.yaml
>>>>> sdhci at 1e130000: $nodename:0: 'sdhci at 1e130000' does not match '^mmc(@.*)?$'
>>>>> From schema: /home/arinc9/Documents/linux/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
>>>>> sdhci at 1e130000: Unevaluated properties are not allowed ('bus-width', 'cap-mmc-highspeed', 'cap-sd-highspeed', 'disable-wp', 'max-frequency', 'vmmc-supply', 'vqmmc-supply' were unexpected)
>>>>> From schema: /home/arinc9/Documents/linux/Documentation/devicetree/bindings/mmc/mtk-sd.yaml
>>>>> xhci at 1e1c0000: $nodename:0: 'xhci at 1e1c0000' does not match '^usb(@.*)?'
>>>>> From schema: /home/arinc9/Documents/linux/Documentation/devicetree/bindings/usb/mediatek,mtk-xhci.yaml
>>>>> xhci at 1e1c0000: compatible: ['mediatek,mt8173-xhci'] is too short
>>>>> From schema: /home/arinc9/Documents/linux/Documentation/devicetree/bindings/usb/mediatek,mtk-xhci.yaml
>>>>> switch0 at 0: $nodename:0: 'switch0 at 0' does not match '^(ethernet-)?switch(@.*)?$'
>>>>> From schema: /home/arinc9/Documents/linux/Documentation/devicetree/bindings/net/dsa/mediatek,mt7530.yaml
>>>>> port at 1: status:0: 'off' is not one of ['okay', 'disabled', 'reserved']
>>>>> From schema: /home/arinc9/.local/lib/python3.10/site-packages/dtschema/schemas/dt-core.yaml
>>>>> port at 2: status:0: 'off' is not one of ['okay', 'disabled', 'reserved']
>>>>> From schema: /home/arinc9/.local/lib/python3.10/site-packages/dtschema/schemas/dt-core.yaml
>>>>> port at 3: status:0: 'off' is not one of ['okay', 'disabled', 'reserved']
>>>>> From schema: /home/arinc9/.local/lib/python3.10/site-packages/dtschema/schemas/dt-core.yaml
>>>>>
>>>>> - Remove "mips,mips1004Kc" compatible string from the cpu nodes. This
>>>>> doesn't exist anywhere.
>>>>> - Change "memc: syscon at 5000" to "memc: memory-controller at 5000".
>>>>> - Change "uartlite: uartlite at c00" to "serial0: serial at c00" and remove the
>>>>> aliases node.
>>>>> - Remove "clock-names" from the serial0 node. The property doesn't exist on
>>>>> the 8250.yaml schema.
>>>>> - Change "sdhci: sdhci at 1e130000" to "mmc: mmc at 1e130000".
>>>>> - Change "xhci: xhci at 1e1c0000" to "usb: usb at 1e1c0000".
>>>>> - Add "mediatek,mtk-xhci" as the second compatible string on the usb node.
>>>>> - Change "switch0: switch0 at 0" to "switch0: switch at 0"
>>>>> - Change "off" to "disabled" for disabled nodes.
>>>>>
>>>>> Remaining warnings are caused by the lack of json-schema documentation.
>>>>>
>>>>> /cpuintc: failed to match any schema with compatible: ['mti,cpu-interrupt-controller']
>>>>> /palmbus at 1e000000/wdt at 100: failed to match any schema with compatible: ['mediatek,mt7621-wdt']
>>>>> /palmbus at 1e000000/i2c at 900: failed to match any schema with compatible: ['mediatek,mt7621-i2c']
>>>>> /palmbus at 1e000000/spi at b00: failed to match any schema with compatible: ['ralink,mt7621-spi']
>>>>> /ethernet at 1e100000: failed to match any schema with compatible: ['mediatek,mt7621-eth']
>>>>>
>>>>> Signed-off-by: Arınç ÜNAL <arinc.unal at arinc9.com>
>>>>> ---
>>>>> .../boot/dts/ralink/mt7621-gnubee-gb-pc1.dts | 2 +-
>>>>> .../boot/dts/ralink/mt7621-gnubee-gb-pc2.dts | 2 +-
>>>>> arch/mips/boot/dts/ralink/mt7621.dtsi | 32 +++++++------------
>>>>> 3 files changed, 14 insertions(+), 22 deletions(-)
>>>>>
>>>>> diff --git a/arch/mips/boot/dts/ralink/mt7621-gnubee-gb-pc1.dts b/arch/mips/boot/dts/ralink/mt7621-gnubee-gb-pc1.dts
>>>>> index 24eebc5a85b1..6ecb8165efe8 100644
>>>>> --- a/arch/mips/boot/dts/ralink/mt7621-gnubee-gb-pc1.dts
>>>>> +++ b/arch/mips/boot/dts/ralink/mt7621-gnubee-gb-pc1.dts
>>>>> @@ -53,7 +53,7 @@ system {
>>>>> };
>>>>> };
>>>>>
>>>>> -&sdhci {
>>>>> +&mmc {
>>>>> status = "okay";
>>>>> };
>>>>>
>>>>> diff --git a/arch/mips/boot/dts/ralink/mt7621-gnubee-gb-pc2.dts b/arch/mips/boot/dts/ralink/mt7621-gnubee-gb-pc2.dts
>>>>> index 34006e667780..2e534ea5bab7 100644
>>>>> --- a/arch/mips/boot/dts/ralink/mt7621-gnubee-gb-pc2.dts
>>>>> +++ b/arch/mips/boot/dts/ralink/mt7621-gnubee-gb-pc2.dts
>>>>> @@ -37,7 +37,7 @@ key-reset {
>>>>> };
>>>>> };
>>>>>
>>>>> -&sdhci {
>>>>> +&mmc {
>>>>> status = "okay";
>>>>> };
>>>>>
>>>>> diff --git a/arch/mips/boot/dts/ralink/mt7621.dtsi b/arch/mips/boot/dts/ralink/mt7621.dtsi
>>>>> index ee46ace0bcc1..9302bdc04510 100644
>>>>> --- a/arch/mips/boot/dts/ralink/mt7621.dtsi
>>>>> +++ b/arch/mips/boot/dts/ralink/mt7621.dtsi
>>>>> @@ -15,13 +15,11 @@ cpus {
>>>>>
>>>>> cpu at 0 {
>>>>> device_type = "cpu";
>>>>> - compatible = "mips,mips1004Kc";
>>>>> reg = <0>;
>>>>> };
>>>>>
>>>>> cpu at 1 {
>>>>> device_type = "cpu";
>>>>> - compatible = "mips,mips1004Kc";
>>>>> reg = <1>;
>>>>> };
>>>>> };
>>>>
>>>> Instead of removing this, since compatible is correct here, I think a
>>>> cpus yaml file needs to be added to properly define mips CPU's but
>>>> compatible strings using all around the sources are a bit messy. Take
>>>> a look of how is this done for arm [0]
>>>
>>> I did investigate the arm bindings beforehand. I've seen that some of
>>> the strings are also checked by code. I don't see the mips strings used
>>> anywhere but DTs so I had decided to remove it here. I guess we can make
>>> a basic binding to list the mips processor cores.
>>
>> At the very least I do think a compatible string should exist for cpu
>> nodes :). And because of the mess with MIPS cpu nodes in dts files all
>> around I think we should only add this 'compatible' as a requirement
>> and mark 'reg' and 'device_type' as optionals.
>
> I have just sent a patch to start from containing all compatible
> strings I have found in the 'arch/mips/boot/dts' folder:
>
> https://lore.kernel.org/linux-devicetree/20220915055514.463241-1-sergio.paracuellos@gmail.com/T/#u
Awesome, I'll keep the string on v2.
>
>>
>>>
>>> What do you think Thomas?
>>>
>>>>
>>>>> @@ -33,11 +31,6 @@ cpuintc: cpuintc {
>>>>> compatible = "mti,cpu-interrupt-controller";
>>>>> };
>>>>>
>>>>> - aliases {
>>>>> - serial0 = &uartlite;
>>>>> - };
>>>>> -
>>>>> -
>>>>> mmc_fixed_3v3: regulator-3v3 {
>>>>> compatible = "regulator-fixed";
>>>>> regulator-name = "mmc_power";
>>>>> @@ -110,17 +103,16 @@ i2c: i2c at 900 {
>>>>> pinctrl-0 = <&i2c_pins>;
>>>>> };
>>>>>
>>>>> - memc: syscon at 5000 {
>>>>> + memc: memory-controller at 5000 {
>>>>> compatible = "mediatek,mt7621-memc", "syscon";
>>>>> reg = <0x5000 0x1000>;
>>>>> };
>>>>>
>>>>
>>>> I think syscon nodes need to use 'syscon' in the node name, but I am
>>>> not 100% sure.
>>>
>>> I've tested this patch series on my GB-PC2, it currently works fine.
>>> Also, DT binding for MT7621 memory controller uses memory-controller on
>>> the example so I guess it's fine?
>>
>> I know that works fine but when the node is a syscon it is good to
>> have that syscon in the node name (I don't know if having it is a rule
>> or something, I guess no). In any case I agree that binding and dts
>> should match.
Understood, I'll keep it syscon in v2.
Arınç
More information about the linux-arm-kernel
mailing list