Re: [PATCH 1/3] dt-bindings: crypto: Convert Atmel AES to yaml

From: Krzysztof Kozlowski
Date: Tue Feb 08 2022 - 06:25:02 EST


On 08/02/2022 10:49, Tudor.Ambarus@xxxxxxxxxxxxx wrote:
> On 2/8/22 10:59, Krzysztof Kozlowski wrote:
>> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
>>
>> On 08/02/2022 05:10, Tudor.Ambarus@xxxxxxxxxxxxx wrote:
>>> Hi, Krzysztof,
>>>
>>> On 2/7/22 17:56, Krzysztof Kozlowski wrote:
>>>> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
>>>>
>>>> On 07/02/2022 04:24, Tudor Ambarus wrote:
>>>>> Convert Atmel AES documentation to yaml format. With the conversion the
>>>>> clock and clock-names properties are made mandatory. The driver returns
>>>>> -EINVAL if "aes_clk" is not found, reflect that in the bindings and make
>>>>> the clock and clock-names properties mandatory. Update the example to
>>>>> better describe how one should define the dt node.
>>>>>
>>>>> Signed-off-by: Tudor Ambarus <tudor.ambarus@xxxxxxxxxxxxx>
>>>>> ---
>>>>> .../devicetree/bindings/crypto/atmel,aes.yaml | 65 +++++++++++++++++++
>>>>> .../bindings/crypto/atmel-crypto.txt | 20 ------
>>>>> 2 files changed, 65 insertions(+), 20 deletions(-)
>>>>> create mode 100644 Documentation/devicetree/bindings/crypto/atmel,aes.yaml
>>>>>
>>>>> diff --git a/Documentation/devicetree/bindings/crypto/atmel,aes.yaml b/Documentation/devicetree/bindings/crypto/atmel,aes.yaml
>>>>> new file mode 100644
>>>>> index 000000000000..f77ec04dbabe
>>>>> --- /dev/null
>>>>> +++ b/Documentation/devicetree/bindings/crypto/atmel,aes.yaml
>>>>> @@ -0,0 +1,65 @@
>>>>> +# SPDX-License-Identifier: GPL-2.0-only
>>>>> +%YAML 1.2
>>>>> +---
>>>>> +$id: http://devicetree.org/schemas/crypto/atmel,aes.yaml#
>>>>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>>>>> +
>>>>> +title: Atmel Advanced Encryption Standard (AES) HW cryptographic accelerator
>>>>> +
>>>>> +maintainers:
>>>>> + - Tudor Ambarus <tudor.ambarus@xxxxxxxxxxxxx>
>>>>> +
>>>>> +properties:
>>>>> + compatible:
>>>>> + const: atmel,at91sam9g46-aes
>>>>> +
>>>>> + reg:
>>>>> + maxItems: 1
>>>>> +
>>>>> + interrupts:
>>>>> + maxItems: 1
>>>>> +
>>>>> + clocks:
>>>>> + maxItems: 1
>>>>> +
>>>>> + clock-names:
>>>>> + const: aes_clk
>>>>> +
>>>>> + dmas:
>>>>> + items:
>>>>> + - description: TX DMA Channel
>>>>> + - description: RX DMA Channel
>>>>> +
>>>>> + dma-names:
>>>>> + items:
>>>>> + - const: tx
>>>>> + - const: rx
>>>>> +
>>>>> +required:
>>>>> + - compatible
>>>>> + - reg
>>>>> + - interrupts
>>>>> + - clocks
>>>>> + - clock-names
>>>>> + - dmas
>>>>> + - dma-names
>>>>> +
>>>>> +additionalProperties: false
>>>>> +
>>>>> +examples:
>>>>> + - |
>>>>> + #include <dt-bindings/interrupt-controller/irq.h>
>>>>> + #include <dt-bindings/interrupt-controller/arm-gic.h>
>>>>> + #include <dt-bindings/clock/at91.h>
>>>>> + #include <dt-bindings/dma/at91.h>
>>>>
>>>> One empty line for readability.
>>>
>>> Ok.
>>>
>>>>
>>>>> + aes: aes@f8038000 {
>>>>
>>>> Generic node name, so "crypto".
>>>
>>> Hm, I'm not convinced why, would you please give more details about this
>>> requirement? This IP is capable of doing just AES operations, I find it
>>> generic enough. We use the "aes" name on all our SoCs that have a version
>>> of this IP, that would be quite a change. So I would prefer to keep the
>>> "aes" name if possible.
>>>
>>
>> The requirement comes from DT specification.
>> "The name of a node should be somewhat generic, reflecting the function
>> of the device and not its precise programming
>> model. If appropriate, the name should be one of the following choice"
>> AES is not generic. AES is specific crypto operation. The spec gives
>> example - "crypto", so use this one just like others are using. Atmel is
>> not special in that matter.
>>
> I see, thanks for the explanation. I will put the node name as "crypto", and add
> a label to it as "aes":
> aes: crypto@f8038000 {
>

That's fine (label naming doesn't matter). You don't have to change all
your DTS files with the new node name, although at some point it could
be required.

For some other bindings in different subsystems, the node name pattern
is now enforced by dtschema (git grep -C 1 -E
'(nodename|patternProperties)' -- Documentation/devicetree/bindings/).

Best regards,
Krzysztof