Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <c5eb5f78-00a7-58d5-4b09-7b6aaf929b56@quicinc.com>
Date:   Tue, 11 Apr 2023 07:46:26 -0700
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:102.0) Gecko/20100101
 Thunderbird/102.9.1
Subject: Re: [PATCH v2 1/2] dt-bindings: firmware: arm,scmi: support parameter
 passing in smc/hvc
Content-Language: en-US
To:     Sudeep Holla <sudeep.holla@arm.com>
CC:     <cristian.marussi@arm.com>, <robh+dt@kernel.org>,
        <krzysztof.kozlowski+dt@linaro.org>,
        <linux-arm-kernel@lists.infradead.org>,
        <devicetree@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
        <lkp@intel.com>
References: <20230409181918.29270-1-quic_nkela@quicinc.com>
 <20230410182058.8949-1-quic_nkela@quicinc.com>
 <20230410182058.8949-2-quic_nkela@quicinc.com>
 <20230411125420.lymhjf5chxdr7if7@bogus>
From:   Nikunj Kela <quic_nkela@quicinc.com>
In-Reply-To: <20230411125420.lymhjf5chxdr7if7@bogus>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk


On 4/11/2023 5:54 AM, Sudeep Holla wrote:
> On Mon, Apr 10, 2023 at 11:20:57AM -0700, Nikunj Kela wrote:
>> Currently, smc/hvc calls are made with smc-id only. The parameters are
>> all set to zeros. This patch defines two optional device tree bindings,
>> that can be used to pass parameters in smc/hvc calls.
>>
> Why 2 values ?

I can do with one property if you prefer that. Its just I wanted to 
ensure that whosoever is setting

the parameter list, is mindful of 32bit vs 64bit convention. If we use 
one property, do you propose to add a new property like width to specify 
if the  parameter list is for 32bit vs 64bit?

>> This is useful when multiple scmi instances are used with common smc-id.
>>
> I really would like to avoid this binding. Because of lack of standard
> SMC/HVC FID for SCMI we had to add this binding. Extending for newer use
> case like this in a vendor specific way is something I would like to avoid.
If you have a solution to get rid of FID from DTB node, I will follow 
the same however until that happens, my view is that we put the 
parameters in dtb.
>
>> Signed-off-by: Nikunj Kela <quic_nkela@quicinc.com>
>> ---
>>   .../devicetree/bindings/firmware/arm,scmi.yaml | 18 ++++++++++++++++++
>>   1 file changed, 18 insertions(+)
>>
>> diff --git a/Documentation/devicetree/bindings/firmware/arm,scmi.yaml b/Documentation/devicetree/bindings/firmware/arm,scmi.yaml
>> index 5824c43e9893..ecf76b763c8c 100644
>> --- a/Documentation/devicetree/bindings/firmware/arm,scmi.yaml
>> +++ b/Documentation/devicetree/bindings/firmware/arm,scmi.yaml
>> @@ -115,6 +115,23 @@ properties:
>>       description:
>>         SMC id required when using smc or hvc transports
>>   
>> +  arm,smc32-params:
>> +    $ref: /schemas/types.yaml#/definitions/uint32-array
>> +    description:
>> +      An optional parameter list passed in smc32 or hvc32 calls
>> +    default: 0
>> +    minItems: 1
>> +    maxItems: 6
>> +
>> +  arm,smc64-params:
>> +    $ref: /schemas/types.yaml#/definitions/uint64-array
>> +    description:
>> +      An optional parameter list passed in smc64 or hvc64 calls.
>> +      This is valid only on ARM64 machines.
>> +    default: 0
>> +    minItems: 1
>> +    maxItems: 6
>> +
> Even if we end up adding(which I would very much like to avoid), I don't see
> the need for 32 and 64 bit params like this. There must be ways to avoid that
> used by some property in some other binding(I will look for one if we choose
> this path)
>