linux-kernel - Re: [PATCH v5 1/3] dt-bindings: media: imx-jpeg: Assign slot for imx jpeg encoder/decoder

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] [day] [month] [year] [list]

Message-ID: <8ce3e155-8787-4a2f-b58e-237afa73fa55@linaro.org>
Date:   Fri, 20 Oct 2023 13:14:28 +0200
From:   Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
To:     ming qian <ming.qian@....nxp.com>, mirela.rabulea@....nxp.com,
        robh+dt@...nel.org, shawnguo@...nel.org
Cc:     krzysztof.kozlowski+dt@...aro.org, conor+dt@...nel.org,
        mchehab@...nel.org, hverkuil-cisco@...all.nl,
        s.hauer@...gutronix.de, kernel@...gutronix.de, festevam@...il.com,
        xiahong.bao@....com, eagle.zhou@....com, tao.jiang_2@....com,
        ming.qian@....com, linux-imx@....com, devicetree@...r.kernel.org,
        linux-media@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH v5 1/3] dt-bindings: media: imx-jpeg: Assign slot for imx
 jpeg encoder/decoder

On 20/10/2023 07:38, ming qian wrote:
> Hi Krzysztof,
> 
>>
>>>> From: Ming Qian <ming.qian@....com>
>>>>
>>>> This IP includes a jpeg wrapper and a jpeg engine, the wrapper is
>>>> working on descriptor based manner. It supports up to 4 slots, each slot
>>>> can have its own chained descriptors. Host won't configure the engine
>>>> directly, but fill some descriptors to encode or decode one jpeg
>>>> picture. Then configure the descriptors to certain slot register. The
>>>> jpeg wrapper will schedule between different slots. When some slot is
>>>> finished, the slot interrupt will be triggered. The purpose of slot is
>>>> that engine can be shared across multiple VMS and os.
>>>>
>>>> Currently, power domains and interrupts are enabled for all 4 slots, but
>>>> only one slot is used. There is no benefit in using more that one slot
>>>> from within the same OS, as the slots are scheduled in round-robin
>>>> manner and not executed in parallel.
>>>>
>>>> Use the property "nxp,slot" to assign a single slot, and just expose the
>>>> parts of the h/w for the assigned slot. For example, only put slot 1's
>>>> power-domains entry in the DT when slot 1 is assigned. If not specified,
>>>> 0 is used by default.
>>>>
>>>> Signed-off-by: Ming Qian <ming.qian@....com>
>>>> ---
>>>> v5
>>>> - improve commit message
>>>> - improve property description
>>>>
>>>> v4
>>>> - improve commit message
>>>> - drop line making the property required, to avoid ABI break
>>>>
>>>> v3
>>>> - add vender prefix, change property slot to nxp,slot
>>>> - add type for property slot
>>>>
>>>> v2
>>>> - add a new property in bindings document
>>>>
>>>>   .../bindings/media/nxp,imx8-jpeg.yaml         | 46 +++++++++----------
>>>>   1 file changed, 22 insertions(+), 24 deletions(-)
>>>>
>>>> diff --git 
>>>> a/Documentation/devicetree/bindings/media/nxp,imx8-jpeg.yaml 
>>>> b/Documentation/devicetree/bindings/media/nxp,imx8-jpeg.yaml
>>>> index 3d9d1db37040..0961856bdcab 100644
>>>> --- a/Documentation/devicetree/bindings/media/nxp,imx8-jpeg.yaml
>>>> +++ b/Documentation/devicetree/bindings/media/nxp,imx8-jpeg.yaml
>>>> @@ -32,19 +32,27 @@ properties:
>>>>       maxItems: 1
>>>>     interrupts:
>>>> -    description: |
>>>> -      There are 4 slots available in the IP, which the driver may use
>>>> -      If a certain slot is used, it should have an associated interrupt
>>>> -      The interrupt with index i is assumed to be for slot i
>>>> -    minItems: 1               # At least one slot is needed by the 
>>>> driver
>>>> -    maxItems: 4               # The IP has 4 slots available for use
>>>> +    description:
>>>> +      Interrupt number for slot
>>>> +    maxItems: 1
>>>
>>> The device still has four interrupts, so we should allow up to four of
>>> them. One given OS might want to use two or all four slots.
>>>
>>>
>> Got it, I will fix it in v6 patch.
>>
> 
> We made an internal discussion about this, current approach of the

How does it help me? Why private discussion should matter to us?

> bindings implies that one dts node is for one slot only, and the slot
> property is integer, not a list of slots. Whoever wants to use 2 or more
> slots, should put more nodes in the dts, one for each slot. Therefor,

Really? Under the same IO address? Please show how it works and then run
`dtbs_check W=1`..

> there is no point in allowing more than one interrupt in one node. Or,
> if we do, for the sake of allowing one OS to use more slots within the
> same dts node, we must also allow more power domains, in any case, we do

power-domains, interrupts, everything.

> not have multiple slots functionality in the driver anymore.

Why your driver should limit bindings? My driver supports multiple slots...

Best regards,
Krzysztof