mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Hans de Goede <hdegoede@redhat.com>
To: Bastien Nocera <hadess@hadess.net>,
	"Luke D. Jones" <luke@ljones.dev>,
	linux-kernel@vger.kernel.org
Cc: pobrn@protonmail.com, linux@roeck-us.net,
	platform-driver-x86@vger.kernel.org
Subject: Re: [PATCH v11] asus-wmi: Add support for custom fan curves
Date: Tue, 28 Sep 2021 13:44:32 +0200	[thread overview]
Message-ID: <ef22aa1f-288a-4e5a-0210-d62c7fc89307@redhat.com> (raw)
In-Reply-To: <1d9d4de26d0c862acf3e579f14de9596027f3c3b.camel@hadess.net>

Hi,

On 9/28/21 1:36 PM, Bastien Nocera wrote:
> On Wed, 2021-09-08 at 11:22 +1200, Luke D. Jones wrote:
>> Add support for custom fan curves found on some ASUS ROG laptops.
>>
>> These laptops have the ability to set a custom curve for the CPU
>> and GPU fans via two ACPI methods.
>>
>> This patch adds two pwm<N> attributes to the hwmon sysfs,
>> pwm1 for CPU fan, pwm2 for GPU fan. Both are under the hwmon of the
>> name `asus_custom_fan_curve`. There is no safety check of the set
>> fan curves - this must be done in userspace.
>>
>> The fans have settings [1,2,3] under pwm<N>_enable:
>> 1. Enable and write settings out
>> 2. Disable and use factory fan mode
>> 3. Same as 2, additionally restoring default factory curve.
>>
>> Use of 2 means that the curve the user has set is still stored and
>> won't be erased, but the laptop will be using its default auto-fan
>> mode. Re-enabling the manual mode then activates the curves again.
>>
>> Notes:
>> - pwm<N>_enable = 0 is an invalid setting.
>> - pwm is actually a percentage and is scaled on writing to device.
> 
> I was trying to update:
> https://gitlab.freedesktop.org/hadess/power-profiles-daemon/-/merge_requests/80
> but I don't understand what files I need to check for what values to
> detect whether custom fan curves were used.
> 
> Can you help me out here?

How to deal with this is actually one of my remaining questions too.

I've not looked at the new code closely yet, but if I understand
things correctly, the now code basically only allows to set 1
custom profile and setting that profile overrides the last
profile set through /sys/firmware/acpi/platform_profile.

And any write to /sys/firmware/acpi/platform_profile will
overwrite / replace the last custom set profile (if any) with
the one matching the requested platform-profile.

So basically users of custom fan profiles are expected to
disable power-profiles-daemon or at least to refrain from
making any platform_profile changes.

And if power-profile-daemon is actually active and
makes a change then any custom settings will be thrown away,
IOW p-p-d will always win. So I believe that it no longer needs
to check for custom profiles, since any time it requests a
standard profile that will overwrite any custom profile
which may be present.

Luke, do I have that right ?

> Also, was this patch accepted in the pdx86 tree?

No, I still need to find/make some time to review it and
I still have the same question as you :)

Regards,

Hans


  parent reply	other threads:[~2021-09-28 11:44 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-07 23:22 [PATCH v11 0/1] " Luke D. Jones
2021-09-07 23:22 ` [PATCH v11] " Luke D. Jones
2021-09-08 12:32   ` kernel test robot
2021-09-08 17:49   ` kernel test robot
2021-09-28 11:36   ` Bastien Nocera
2021-09-28 11:43     ` Luke Jones
2021-09-28 11:56       ` Hans de Goede
2021-09-28 11:59         ` Luke Jones
2021-09-28 12:03           ` Hans de Goede
2021-09-28 12:15             ` Luke Jones
2021-09-28 14:11               ` Hans de Goede
2021-09-28 11:44     ` Hans de Goede [this message]
2021-09-28 11:56       ` Luke Jones
2021-09-28 11:59         ` Hans de Goede
2021-09-28 12:01           ` Luke Jones

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=ef22aa1f-288a-4e5a-0210-d62c7fc89307@redhat.com \
    --to=hdegoede@redhat.com \
    --cc=hadess@hadess.net \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux@roeck-us.net \
    --cc=luke@ljones.dev \
    --cc=platform-driver-x86@vger.kernel.org \
    --cc=pobrn@protonmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®