linux-kernel - Re: [PATCH 1/2] platform/x86: dell-ddv: Add hwmon support

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]

Message-ID: <08af724c-8c65-db41-ece8-f330f67a3dc2@gmx.de>
Date:   Mon, 6 Feb 2023 22:19:47 +0100
From:   Armin Wolf <W_Armin@....de>
To:     Guenter Roeck <linux@...ck-us.net>,
        Hans de Goede <hdegoede@...hat.com>
Cc:     markgross@...nel.org, jdelvare@...e.com,
        platform-driver-x86@...r.kernel.org, linux-hwmon@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/2] platform/x86: dell-ddv: Add hwmon support

Am 06.02.23 um 17:23 schrieb Guenter Roeck:

> On Mon, Feb 06, 2023 at 03:13:25PM +0100, Hans de Goede wrote:
>> Hi Armin,
>>
>> On 2/5/23 21:54, Armin Wolf wrote:
>>> Thanks to bugreport 216655 on bugzilla triggered by the
>>> dell-smm-hwmon driver, the contents of the sensor buffers
>>> could be almost completely decoded.
>>> Add an hwmon interface for exposing the fan and thermal
>>> sensor values. Since the WMI interface can be quite slow
>>> on some machines, the sensor buffers are cached for 1 second
>>> to lessen the performance impact.
>>> The debugfs interface remains in place to aid in reverse-engineering
>>> of unknown sensor types and the thermal buffer.
>>>
>>> Tested-by: Antonín Skala <skala.antonin@...il.com>
>>> Tested-by: Gustavo Walbon <gustavowalbon@...il.com>
>>> Signed-off-by: Armin Wolf <W_Armin@....de>
>> This looks nice and clean to me, thank you:
>>
>> Reviewed-by: Hans de Goede <hdegoede@...hat.com>
>>
>> I'm going to wait a bit with merging this to see if Guenter
>> has any remarks. If there are no remarks by next Monday then
>> I'll merge this for the 6.3 merge window.
>>
> I don't have any further remarks. I won't send an Ack, though.
> I noticed that the empty lines before return statements distract
> me sufficiently enough that I am not sure about the actual code.
>
> Guenter

I will send a 3rd revision of the patch series to remove those empty lines then.
Speaking of a 3rd revision, i noticed that jiffies are not updated during suspend,
so the sensor values might be wrong for a short time when resuming after a long time.

This can be solved by either usingktime_get_boottime_ns(), which does not stop during suspend, or by
invalidating the sensor cache upon resume. Which method should be used
in this case? Armin Wolf