linux-kernel - Re: [PATCH] dt-bindings: mailbox: mtk-gce: Convert txt to json-schema

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 24 May 2022 14:18:38 +0200
From:   AngeloGioacchino Del Regno 
        <angelogioacchino.delregno@...labora.com>
To:     Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>,
        robh+dt@...nel.org
Cc:     jassisinghbrar@...il.com, krzysztof.kozlowski+dt@...aro.org,
        matthias.bgg@...il.com, houlong.wei@...iatek.com,
        linux-kernel@...r.kernel.org, devicetree@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org,
        linux-mediatek@...ts.infradead.org
Subject: Re: [PATCH] dt-bindings: mailbox: mtk-gce: Convert txt to json-schema

Il 20/05/22 10:15, Krzysztof Kozlowski ha scritto:
> On 19/05/2022 12:18, AngeloGioacchino Del Regno wrote:
>> Convert the mtk-gce documentation from freeform text format to a
>> json-schema.
>>
>> Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@...labora.com>
>> ---
>>   .../bindings/mailbox/mediatek,gce-mbox.yaml   | 114 ++++++++++++++++++
>>   .../devicetree/bindings/mailbox/mtk-gce.txt   |  82 -------------
>>   2 files changed, 114 insertions(+), 82 deletions(-)
>>   create mode 100644 Documentation/devicetree/bindings/mailbox/mediatek,gce-mbox.yaml
>>   delete mode 100644 Documentation/devicetree/bindings/mailbox/mtk-gce.txt
>>
>> diff --git a/Documentation/devicetree/bindings/mailbox/mediatek,gce-mbox.yaml b/Documentation/devicetree/bindings/mailbox/mediatek,gce-mbox.yaml
>> new file mode 100644
>> index 000000000000..750391b4038c
>> --- /dev/null
>> +++ b/Documentation/devicetree/bindings/mailbox/mediatek,gce-mbox.yaml
>> @@ -0,0 +1,114 @@
>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
>> +%YAML 1.2
>> +---
>> +$id: http://devicetree.org/schemas/mailbox/mediatek,gce-mbox.yaml#
>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>> +
>> +title: MediaTek Global Command Engine (GCE) mailbox
>> +
>> +maintainers:
>> +  - Houlong Wei <houlong.wei@...iatek.com>
>> +
>> +description: |
>> +  The Global Command Engine (GCE) is used to help read/write registers
>> +  with critical time limitation, such as updating display configuration
>> +  during the vblank.
>> +  The GCE can be used to implement the Command Queue (CMDQ) driver.
> 
> Mention the headers in description.
> 

   Values for properties used by the GCE, such as sub-system IDs, thread

   priority and event IDs are defined in 'dt-bindings/gce/<chip>-gce.h'.

Would that be enough, or should I list all of the headers?

>> +
>> +properties:
>> +  compatible:
>> +    enum:
>> +      - mediatek,mt6779-gce
>> +      - mediatek,mt8173-gce
>> +      - mediatek,mt8183-gce
>> +      - mediatek,mt8186-gce
>> +      - mediatek,mt8192-gce
>> +      - mediatek,mt8195-gce
>> +
>> +  reg:
>> +    maxItems: 1
>> +
>> +  interrupts:
>> +    maxItems: 1
>> +
>> +  clocks:
>> +    maxItems: 1
>> +
>> +  clock-names:
>> +    items:
>> +      - const: gce
>> +
>> +  '#mbox-cells':
>> +    description: |
>> +      The first cell describes the mailbox channel, which is the GCE Thread ID;
>> +      The second cell describes the priority of the GCE thread.
>> +    const: 2
>> +
>> +required:
>> +  - compatible
>> +  - reg
>> +  - interrupts
>> +  - clocks
>> +  - clock-names
>> +  - '#mbox-cells'
>> +
>> +additionalProperties: false
>> +
>> +allOf:
>> +  - if:
>> +      properties:
>> +        compatible:
>> +          enum:
>> +            - mediatek,mt8195-gce
>> +    then:
>> +      properties:
>> +        clocks:
>> +          maxItems: 2
> 
> Are you sure this works on mt8195-gce?
> 

Thanks for that, I've just rechecked the driver and.. no, that won't
work for MT8195: it's just one clock there (like the others) and the
clock names aren't even enforced, as the driver is always taking the
clock at index 0.

I got confused because the driver uses a slightly different kind of
logic when probing on SoCs with multiple mailboxes, specifically:
  - For single mailbox, having a clock with name "gce" is enforced
    as it's grabbing it with devm_clk_get(dev, clk_name), where the
    clock name is declared in a string called "clk_name";
  - For multiple mailboxes, it's looking for an of_alias, declared
    in an array of strings called "clk_names" and getting the clock
    with of_clk_get(node, 0).


So there comes my confusion, recapping:

static const char * const clk_name = "gce";
                  <- this is a clock name
static const char * const clk_names[] = { "gce0", "gce1" }; <- OF alias names


At this point, I think that the best idea would be to fix this issue
first... luckily there's no MT8195 devicetree upstream yet, so I would
technically not be breaking any ABI by changing it to be the same as
the others.

Easier explanation: plan is to change the driver such that we won't
need anything different from the others in this schema.

New version coming soon, then....

Regards,
Angelo