linux-kernel - Re: [PATCH 1/6] thermal: hwmon: Replace the call the thermal_cdev

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]

Date:   Sat, 11 Apr 2020 23:07:20 +0200
From:   Daniel Lezcano <daniel.lezcano@...aro.org>
To:     Guenter Roeck <linux@...ck-us.net>, rui.zhang@...el.com
Cc:     amit.kucheria@...durent.com, linux-kernel@...r.kernel.org,
        Kamil Debski <kamil@...as.org>,
        Bartlomiej Zolnierkiewicz <b.zolnierkie@...sung.com>,
        Jean Delvare <jdelvare@...e.com>,
        "open list:PWM FAN DRIVER" <linux-hwmon@...r.kernel.org>,
        Lukasz Majewski <lukma@...x.de>
Subject: Re: [PATCH 1/6] thermal: hwmon: Replace the call the
 thermal_cdev_update()

On 11/04/2020 19:26, Guenter Roeck wrote:
> On 4/11/20 9:45 AM, Daniel Lezcano wrote:
>> On 11/04/2020 03:32, Guenter Roeck wrote:
>>> On 4/10/20 3:12 PM, Daniel Lezcano wrote:
>>>> The function thermal_cdev_upadte is called from the throttling
>>>
>>> misspelled
>>>
>>>> functions in the governors not from the cooling device itself.
>>>>
>>>> The cooling device is set to its maximum state and then updated. Even
>>>> if I don't get the purpose of probing the pwm-fan to its maximum
>>>> cooling state, we can replace the thermal_cdev_update() call to the
>>>> internal set_cur_state() function directly.
>>>>
>>>> Signed-off-by: Daniel Lezcano <daniel.lezcano@...aro.org>
>>>> ---
>>>>  drivers/hwmon/pwm-fan.c | 3 +--
>>>>  1 file changed, 1 insertion(+), 2 deletions(-)
>>>>
>>>> diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c
>>>> index 30b7b3ea8836..a654ecdf21ab 100644
>>>> --- a/drivers/hwmon/pwm-fan.c
>>>> +++ b/drivers/hwmon/pwm-fan.c
>>>> @@ -372,7 +372,6 @@ static int pwm_fan_probe(struct platform_device *pdev)
>>>>  	if (ret)
>>>>  		return ret;
>>>>  
>>>> -	ctx->pwm_fan_state = ctx->pwm_fan_max_state;
>>>>  	if (IS_ENABLED(CONFIG_THERMAL)) {
>>>>  		cdev = devm_thermal_of_cooling_device_register(dev,
>>>>  			dev->of_node, "pwm-fan", ctx, &pwm_fan_cooling_ops);
>>>> @@ -384,7 +383,7 @@ static int pwm_fan_probe(struct platform_device *pdev)
>>>>  			return ret;
>>>>  		}
>>>>  		ctx->cdev = cdev;
>>>> -		thermal_cdev_update(cdev);
>>>> +		pwm_fan_set_cur_state(cdev, ctx->pwm_fan_max_state);
>>>
>>> So far the function would only change the state if the new
>>> state is not equal to the old state. This was the case because
>>> pwm_fan_state was set to pwm_fan_max_state, and the call to
>>> thermal_cdev_update() and thus pwm_fan_set_cur_state() would
>>> do nothing except update statistics. The old code _assumed_
>>> that the current state is pwm_fan_max_state. The new code
>>> enforces it. That is a substantial semantic change, and it
>>> is not really reflected in the commit message. Is that really
>>> what you want ? If so, the commit message needs to state that
>>> and explain the rationale.
>>
>> Well, to be honest I'm not getting the rational of calling
>> thermal_cdev_update(cdev) right after
>> devm_thermal_of_cooling_device_register() neither setting pwm_fan_state
>> to pwm_fan_max_state.
>>
> Good question. The author might know/recall. Maybe the idea was that
> thermal would update the state to a lower state shortly thereafter.
> 
>> Do we have the guarantee there is at this point a thermal instance
>> making the target state working when thermal_cdev_update is called?
>>
>> Are we sure a thermal_cdev_update(cdev) is actually right here?
>>
> I don't know. I am not exactly familiar with thermal subsystem
> particulars. I do recall seeing similar code in other drivers, though.

This call is done only in the governors actually.

> Either case, your patch does change functionality, and we should not
> do that without understanding its impact.

Right, so I've been hacking my board, added a pwm-fan and binded the
thermal zone to it.

As expected, the call to thermal_cdev_update() is not needed.

ctx->pwm_fan_state = ctx->pwm_fan_max_state;

intializes to a max value (in my case it is 3). Right after it calls
thermal_cdev_update() which fails to find any instance active because we
are at init time and then calls set_cur_state with the target state set
to zero and passing through a stats usage for nothing.

The ctx->pwm_fan_state is only used by the cooling device ops, so I
don't see any reason why it is set to pwm_fan_max_state before the
compilation condition.

May be there is something subtle here.

Lukasz ? Is there any reason why thermal_cdev_update() was called here ?

IMO, this function is a governor thing and it must be removed from the
cooling device.



-- 
<http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs

Follow Linaro:  <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog