linux-kernel - Re: [PATCH v2 3/3] dt-bindings: hwmon: max31790: Add maxim,pwmout-pin-as-tach-input property

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: Sun, 5 May 2024 08:40:39 -0700
From: Guenter Roeck <linux@...ck-us.net>
To: Chanh Nguyen <chanh@...eremail.onmicrosoft.com>,
 Conor Dooley <conor@...nel.org>
Cc: Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>,
 Chanh Nguyen <chanh@...amperecomputing.com>, Jean Delvare
 <jdelvare@...e.com>, Rob Herring <robh+dt@...nel.org>,
 Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
 Conor Dooley <conor+dt@...nel.org>, Justin Ledford
 <justinledford@...gle.com>, devicetree@...r.kernel.org,
 linux-hwmon@...r.kernel.org, linux-kernel@...r.kernel.org,
 OpenBMC Maillist <openbmc@...ts.ozlabs.org>,
 Open Source Submission <patches@...erecomputing.com>,
 Phong Vo <phong@...amperecomputing.com>,
 Thang Nguyen <thang@...amperecomputing.com>,
 Quan Nguyen <quan@...amperecomputing.com>
Subject: Re: [PATCH v2 3/3] dt-bindings: hwmon: max31790: Add
 maxim,pwmout-pin-as-tach-input property

On 5/5/24 03:08, Chanh Nguyen wrote:
> 
> 
> On 25/04/2024 21:05, Guenter Roeck wrote:
>> On 4/25/24 03:33, Chanh Nguyen wrote:
>>>
>>>
>>> On 24/04/2024 00:02, Conor Dooley wrote:
>>>> [EXTERNAL EMAIL NOTICE: This email originated from an external sender. Please be mindful of safe email handling and proprietary information protection practices.]
>>>>
>>>
>>
>> The quote doesn't make much sense.
>>
>>> Sorry Conor, there may be confusion here. I mean the mapping of the PWM output to the TACH input, which is on the MAX31790, and it is not sure a common feature on all fan controllers.
>>>
>>
>> I think the term "mapping" is a bit confusing here.
>>
>> tach-ch, as I understand it, is supposed to associate a tachometer input
>> with a pwm output, meaning the fan speed measured with the tachometer input
>> is expected to change if the pwm output changes.
>>
>> On MAX31790, it is possible to configure a pwm output pin as tachometer input pin.
>> That is something completely different. Also, the association is fixed.
>> If the first pwm channel is used as tachometer channel, it would show up as 7th
>> tachometer channel. If the 6th pwm channel is configured to be used as tachometer
>> input, it would show up as 12th tachometer channel.
>>
>> Overall, the total number of channels on MAX31790 is always 12. 6 of them
>> are always tachometer inputs, the others can be configured to either be a
>> pwm output or a tachometer input.
> 
> Thank you, Guenter, for your explanation. That is also my understanding of the MAX31790 feature.
> 
> So, I think we should introduce a vendor property to configure the pwm output pins to become tachometer input pins. We shouldn't use the tach-ch property. Because they are completely different, I think.
> 
> What's your idea ? Please help share me, Guenter
> 
> 
>>
>> pwm outputs on MAX31790 are always tied to the matching tachometer inputs
>> (pwm1 <--> tach1 etc) and can not be reconfigured, meaning tach-ch for
>> channel X would always be X.
>>
>>> I would like to open a discussion about whether we should use the tach-ch property on the fan-common.yaml
>>>
>>> I'm looking forward to hearing comments from everyone. For me, both tach-ch and vendor property are good.
>>>
>>
>> I am not even sure how to define tach-ch to mean "use the pwm output pin
>> associated with this tachometer input channel not as pwm output
>> but as tachometer input". That would be a boolean, not a number.
>>
> 
> Thank Guenter,
> 
> I reviewed again the "tach-ch" property, which is used in the https://elixir.bootlin.com/linux/v6.9-rc6/source/Documentation/devicetree/bindings/hwmon/aspeed,g6-pwm-tach.yaml#L68 and https://elixir.bootlin.com/linux/v6.9-rc6/source/drivers/hwmon/aspeed-g6-pwm-tach.c#L434
> 
> That is something completely different from my purpose.
> 

Based on its definition, tach-ch is associated with fans, and it looks
like the .yaml file groups multiple sets of fans into a single
fan node.

In the simple case that would be
	tach-ch = <1>
..
	tach-ch = <12>

or, if all fans are controlled by a single pwm
	tach-ch = <1 2 3 4 5 6 8 9 10 11 12>

The existence of tachometer channel 7..12 implies that pwm channel (tachometer
channel - 6) is used as tachometer channel. That should be sufficient to program
the chip for that channel. All you'd have to do is to ensure that pwm channel
"X" is not listed as tachometer channel "X + 6", and program pwm channel "X - 6"
for tachometer channels 7..12 as tachometer channels.

Hope this helps,
Guenter