Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [v4 07/11] dt-bindings: hwmon: Add documents for PECI hwmon
 client drivers
To:     Rob Herring <robh@kernel.org>
Cc:     Mark Rutland <mark.rutland@arm.com>,
        Haiyue Wang <haiyue.wang@linux.intel.com>,
        Vernon Mauery <vernon.mauery@linux.intel.com>,
        James Feist <james.feist@linux.intel.com>,
        devicetree@vger.kernel.org,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Andrew Jeffery <andrew@aj.id.au>,
        Arnd Bergmann <arnd@arndb.de>,
        Jason M Biils <jason.m.bills@linux.intel.com>,
        Joel Stanley <joel@jms.id.au>
References: <20180521195905.27983-1-jae.hyun.yoo@linux.intel.com>
 <20180522164230.GA11707@rob-hp-laptop>
 <a05b44a2-ecc8-b028-4308-47d88523a6f1@linux.intel.com>
 <CAL_JsqJ8rjRzaqasjbdq-jkYfWxdo72YnmKAVNxDwo6wtyv1xg@mail.gmail.com>
From:   Jae Hyun Yoo <jae.hyun.yoo@linux.intel.com>
Message-ID: <51752610-d2c1-7fbc-101e-e99346fa29e4@linux.intel.com>
Date:   Wed, 23 May 2018 09:37:59 -0700
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:52.0) Gecko/20100101
 Thunderbird/52.6.0
MIME-Version: 1.0
In-Reply-To: <CAL_JsqJ8rjRzaqasjbdq-jkYfWxdo72YnmKAVNxDwo6wtyv1xg@mail.gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 5/23/2018 8:11 AM, Rob Herring wrote:
> On Tue, May 22, 2018 at 12:18 PM, Jae Hyun Yoo
> <jae.hyun.yoo@linux.intel.com> wrote:
>> On 5/22/2018 9:42 AM, Rob Herring wrote:
>>>
>>> On Mon, May 21, 2018 at 12:59:05PM -0700, Jae Hyun Yoo wrote:
>>>>
>>>> This commit adds dt-bindings documents for PECI hwmon client drivers.
>>>>
>>>> Signed-off-by: Jae Hyun Yoo <jae.hyun.yoo@linux.intel.com>
>>>> Reviewed-by: Haiyue Wang <haiyue.wang@linux.intel.com>
>>>> Reviewed-by: James Feist <james.feist@linux.intel.com>
>>>> Reviewed-by: Vernon Mauery <vernon.mauery@linux.intel.com>
>>>> Cc: Andrew Jeffery <andrew@aj.id.au>
>>>> Cc: Arnd Bergmann <arnd@arndb.de>
>>>> Cc: Jason M Biils <jason.m.bills@linux.intel.com>
>>>> Cc: Joel Stanley <joel@jms.id.au>
>>>> ---
>>>>    .../bindings/hwmon/peci-cputemp.txt           | 23 ++++++++++++++++++
>>>>    .../bindings/hwmon/peci-dimmtemp.txt          | 24 +++++++++++++++++++
>>>>    2 files changed, 47 insertions(+)
>>>>    create mode 100644
>>>> Documentation/devicetree/bindings/hwmon/peci-cputemp.txt
>>>>    create mode 100644
>>>> Documentation/devicetree/bindings/hwmon/peci-dimmtemp.txt
>>>>
>>>> diff --git a/Documentation/devicetree/bindings/hwmon/peci-cputemp.txt
>>>> b/Documentation/devicetree/bindings/hwmon/peci-cputemp.txt
>>>> new file mode 100644
>>>> index 000000000000..2f59aee12d9e
>>>> --- /dev/null
>>>> +++ b/Documentation/devicetree/bindings/hwmon/peci-cputemp.txt
>>>> @@ -0,0 +1,23 @@
>>>> +Bindings for Intel PECI (Platform Environment Control Interface) cputemp
>>>> driver.
>>>> +
>>>> +Required properties:
>>>> +- compatible : Should be "intel,peci-cputemp".
>>>> +- reg        : Should contain address of a client CPU. Address range of
>>>> CPU
>>>> +              clients is starting from 0x30 based on PECI specification.
>>>> +
>>>> +Example:
>>>> +       peci-bus@0 {
>>>> +               #address-cells = <1>;
>>>> +               #size-cells = <0>;
>>>> +               < more properties >
>>>> +
>>>> +               peci-cputemp@30 {
>>>> +                       compatible = "intel,peci-cputemp";
>>>> +                       reg = <0x30>;
>>>> +               };
>>>
>>>
>>> [...]
>>>
>>>> +               peci-dimmtemp@30 {
>>>> +                       compatible = "intel,peci-dimmtemp";
>>>> +                       reg = <0x30>;
>>>> +               };
>>>
>>>
>>> As I said in the prior version, 2 nodes at the same address is wrong.
>>>
>>> Rob
>>>
>>
>> In PECI bus, there is one and only bus host (adapter) and multiple
>> clients on a PECI bus, and PECI spec doesn't allow multiple originators
>> so only the host device can originate message.
> 
> Yes, I get that. A single host still has to address slave devices.
> 
>> In this implementation,
>> all message transactions on a bus from client driver modules and user
>> space will be serialized well in the PECI core bus driver so bus
>> occupation and traffic arbitration will be managed well in the PECI core
>> bus driver even in case of a bus has 2 client drivers at the same
>> address. I'm sure that this implementation doesn't make that kind of
>> problem to OS.
> 
> Multiple clients to a single device is common, but that is a software
> problem and doesn't belong in DT.
> 
> I don't think there is a single other case in the kernel where
> multiple drivers can bind to the same device at a given bus address.
> That is why we have things like MFD. Though in this case, why can't
> one hwmon driver register multiple hwmon devices (cpu and dimm temps)?
> 

It was implemented as a single driver until v2 but dimm temps need
delayed creation unlikely the cpu temps on hwmon subsystem because of
memory training behavior of remote x86 cpus. Since hwmon doesn't allow
incremental creation, I had to divide it into two, cputemp and dimmtemp,
so that cputemp can be registered immediately when the remote x86 cpu
turns on and dimmtemp can be registered by delayed creation. It is the
reason why I had to make the two hwmon driver modules that sharing a
single device address. Additionally, PECI isn't limited for temperature
monitoring feature but it can be used for other functions such as
platform management, cpu interface tuning and diagnostics and failure
analysis, so in case of adding a new driver for the functions, we should
add an another DT node which is sharing the same cpu address.

Thanks,

-Jae