[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <f3e3a3d0-6d21-c782-38a2-c8b2c36242c3@linaro.org>
Date: Thu, 8 Dec 2022 11:27:27 +0100
From: Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
To: Miquel Raynal <miquel.raynal@...tlin.com>
Cc: Xiangsheng Hou <xiangsheng.hou@...iatek.com>,
Richard Weinberger <richard@....at>,
Vignesh Raghavendra <vigneshr@...com>,
Rob Herring <robh+dt@...nel.org>,
Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
Matthias Brugger <matthias.bgg@...il.com>,
Mark Brown <broonie@...nel.org>,
Chuanhong Guo <gch981213@...il.com>,
linux-mtd@...ts.infradead.org, devicetree@...r.kernel.org,
linux-arm-kernel@...ts.infradead.org,
linux-mediatek@...ts.infradead.org, linux-kernel@...r.kernel.org,
linux-spi@...r.kernel.org, benliang.zhao@...iatek.com,
bin.zhang@...iatek.com
Subject: Re: [PATCH v3 7/9] dt-bindings: mtd: Split ECC engine with rawnand
controller
On 08/12/2022 11:00, Miquel Raynal wrote:
> Hi Krzysztof,
>
> krzysztof.kozlowski@...aro.org wrote on Thu, 8 Dec 2022 10:44:17 +0100:
>
>> On 08/12/2022 07:29, Xiangsheng Hou wrote:
>>> Split MediaTek ECC engine with rawnand controller and convert to
>>> YAML schema.
>>>
>>> Signed-off-by: Xiangsheng Hou <xiangsheng.hou@...iatek.com>
>>> ---
>>> .../bindings/mtd/mediatek,mtk-nfc.yaml | 154 +++++++++++++++
>>> .../mtd/mediatek,nand-ecc-engine.yaml | 62 ++++++
>>> .../devicetree/bindings/mtd/mtk-nand.txt | 176 ------------------
>>> 3 files changed, 216 insertions(+), 176 deletions(-)
>>> create mode 100644 Documentation/devicetree/bindings/mtd/mediatek,mtk-nfc.yaml
>>> create mode 100644 Documentation/devicetree/bindings/mtd/mediatek,nand-ecc-engine.yaml
>>> delete mode 100644 Documentation/devicetree/bindings/mtd/mtk-nand.txt
>>>
>>> diff --git a/Documentation/devicetree/bindings/mtd/mediatek,mtk-nfc.yaml b/Documentation/devicetree/bindings/mtd/mediatek,mtk-nfc.yaml
>>> new file mode 100644
>>> index 000000000000..eb1a44c7ae4e
>>> --- /dev/null
>>> +++ b/Documentation/devicetree/bindings/mtd/mediatek,mtk-nfc.yaml
>>> @@ -0,0 +1,154 @@
>>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
>>> +%YAML 1.2
>>> +---
>>> +$id: http://devicetree.org/schemas/mtd/mediatek,mtk-nfc.yaml#
>>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>>> +
>>> +title: MediaTek(MTK) SoCs raw NAND FLASH controller (NFC)
>>> +
>>> +maintainers:
>>> + - Xiangsheng Hou <xiangsheng.hou@...iatek.com>
>>> +
>>> +properties:
>>> + compatible:
>>> + enum:
>>> + - mediatek,mt2701-nfc
>>> + - mediatek,mt2712-nfc
>>> + - mediatek,mt7622-nfc
>>> +
>>> + reg:
>>> + items:
>>> + - description: Base physical address and size of NFI.
>>> +
>>> + interrupts:
>>> + items:
>>> + - description: NFI interrupt
>>> +
>>> + clocks:
>>> + items:
>>> + - description: clock used for the controller
>>> + - description: clock used for the pad
>>> +
>>> + clock-names:
>>> + items:
>>> + - const: nfi_clk
>>> + - const: pad_clk
>>> +
>>> + ecc-engine:
>>> + description: device-tree node of the required ECC engine.
>>> + $ref: /schemas/types.yaml#/definitions/phandle
>>> +
>>> +patternProperties:
>>> + "^nand@[a-f0-9]$":
>>> + type: object
>>
>> This should be instead:
>> $ref: nand-chip.yaml#
>> unevaluatedProperties: false
>>
>> and then properties below (due to current dtschema limitations) should
>> list properties from nand-controller.yaml:
>>
>> nand-on-flash-bbt: true
>>
>> Optionally, we could create additional schema - nand-controller-chip,
>> which would be referenced directly by nand-controller and itself would
>> ref nand-chip.
>
> Isn't this enough? (in linux-next)
> https://git.kernel.org/pub/scm/linux/kernel/git/mtd/linux.git/tree/Documentation/devicetree/bindings/mtd/nand-controller.yaml?h=mtd/next#n54
No, I tested it and it does not work as intended. In this particular
case. I think this is a limitation of dtschema, because binding itself
looks fine. The problem is that you have:
1. mtk-nfc having nand@ children. mtk-nfc references nand-controller
which brings these children.
2. However nand-controller while bringing these children does two things:
a. ref: nand-chip
b. add more propeties
3. The mtk-nfc must further extend the nand@ child.
4. If you add "unevaluatedProperties: false" you notice warnings of
unevaluated propertie from nand-controller children.
Best regards,
Krzysztof
Powered by blists - more mailing lists