From: Conor Dooley <conor.dooley@microchip.com>

The commit in the fixes tag removed the clock-names property from the
SiFive FU740 PCI Controller dt-binding, but it was already in the dts
for the FU740. dtbs_check was not able to pick up on this at the time
but v2022.08 of dt-schema now can:

arch/riscv/boot/dts/sifive/hifive-unmatched-a00.dtb: pcie@e00000000: Unevaluated properties are not allowed ('clock-names' was unexpected)
        From schema: linux/Documentation/devicetree/bindings/pci/sifive,fu740-pcie.yaml

The Linux driver does not use this property, but outside of the kernel
this property may have users. Re-add the property and its "clocks"
dependency.

Fixes: b92225b034c0 ("dt-bindings: PCI: designware: Fix 'unevaluatedProperties' warnings")
Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
---
I went back and forth on removing the property from the dts, but this
seems like the change that is more conservative..
---
 .../devicetree/bindings/pci/sifive,fu740-pcie.yaml          | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/Documentation/devicetree/bindings/pci/sifive,fu740-pcie.yaml b/Documentation/devicetree/bindings/pci/sifive,fu740-pcie.yaml
index 195e6afeb169..c7a9a2dc0fa6 100644
--- a/Documentation/devicetree/bindings/pci/sifive,fu740-pcie.yaml
+++ b/Documentation/devicetree/bindings/pci/sifive,fu740-pcie.yaml
@@ -51,6 +51,12 @@ properties:
     description: A phandle to the PCIe power up reset line.
     maxItems: 1
 
+  clocks:
+    maxItems: 1
+
+  clock-names:
+    const: pcie_aux
+
   pwren-gpios:
     description: Should specify the GPIO for controlling the PCI bus device power on.
     maxItems: 1
-- 
2.37.1

On 11/08/2022 23:33, Conor Dooley wrote:
> From: Conor Dooley <conor.dooley@microchip.com>
> 
> The commit in the fixes tag removed the clock-names property from the

Instead:
The commit b92225b034c0 ("dt-bindings: PCI: designware: Fix
'unevaluatedProperties' warnings")....

> SiFive FU740 PCI Controller dt-binding,

No, it did not do it... At least I cannot see it. Where is the removal
exactly in that patch? The commit removed clock-names from required, not
from properties.

 but it was already in the dts
> for the FU740. dtbs_check was not able to pick up on this at the time
> but v2022.08 of dt-schema now can:
> 
> arch/riscv/boot/dts/sifive/hifive-unmatched-a00.dtb: pcie@e00000000: Unevaluated properties are not allowed ('clock-names' was unexpected)
>         From schema: linux/Documentation/devicetree/bindings/pci/sifive,fu740-pcie.yaml
> 
> The Linux driver does not use this property, but outside of the kernel
> this property may have users. Re-add the property and its "clocks"
> dependency.
> 


Best regards,
Krzysztof

On 12/08/2022 08:34, Krzysztof Kozlowski wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> 
> On 11/08/2022 23:33, Conor Dooley wrote:
>> From: Conor Dooley <conor.dooley@microchip.com>
>>
>> The commit in the fixes tag removed the clock-names property from the
> 
> Instead:
> The commit b92225b034c0 ("dt-bindings: PCI: designware: Fix
> 'unevaluatedProperties' warnings")....
> 
>> SiFive FU740 PCI Controller dt-binding,
> 
> No, it did not do it... At least I cannot see it. Where is the removal
> exactly in that patch? The commit removed clock-names from required, not
> from properties.

I rewrote the commmit message 3 times as I went back and forth on the
"right" change to make, prob just added the wrong fixes tag to a copy
paste of the original message.
Thanks,
Conor.

> 
>   but it was already in the dts
>> for the FU740. dtbs_check was not able to pick up on this at the time
>> but v2022.08 of dt-schema now can:
>>
>> arch/riscv/boot/dts/sifive/hifive-unmatched-a00.dtb: pcie@e00000000: Unevaluated properties are not allowed ('clock-names' was unexpected)
>>          From schema: linux/Documentation/devicetree/bindings/pci/sifive,fu740-pcie.yaml
>>
>> The Linux driver does not use this property, but outside of the kernel
>> this property may have users. Re-add the property and its "clocks"
>> dependency.
>>
> 
> 
> Best regards,
> Krzysztof