Re: [PATCH v6 09/13] HID: i2c-hid: acpi: Stop setting wakeup_capable

[Limonciello, Mario]

On 1/12/2023 13:57, Werner Sembach wrote:
> Hi,
>
> Am 29.09.22 um 18:19 schrieb Raul E Rangel:
> > This is now handled by the i2c-core drive >
> > Signed-off-by: Raul E Rangel <rrangel@xxxxxxxxxxxx>
> > Acked-by: Benjamin Tissoires <benjamin.tissoires@xxxxxxxxxx>
> > ---
> >
> > (no changes since v5)
> >
> > Changes in v5:
> > - Added Acked-by: Benjamin Tissoires
> >
> >   drivers/hid/i2c-hid/i2c-hid-acpi.c | 5 -----
> >   1 file changed, 5 deletions(-)
> >
> > diff --git a/drivers/hid/i2c-hid/i2c-hid-acpi.c 
> > b/drivers/hid/i2c-hid/i2c-hid-acpi.c
> > index b96ae15e0ad917e..375c77c3db74d92 100644
> > --- a/drivers/hid/i2c-hid/i2c-hid-acpi.c
> > +++ b/drivers/hid/i2c-hid/i2c-hid-acpi.c
> > @@ -105,11 +105,6 @@ static int i2c_hid_acpi_probe(struct i2c_client 
> > *client)
> >       acpi_device_fix_up_power(adev);
> > -    if (acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0) {
> > -        device_set_wakeup_capable(dev, true);
> > -        device_set_wakeup_enable(dev, false);
> > -    }
> > -
> >       return i2c_hid_core_probe(client, &ihid_acpi->ops,
> >                     hid_descriptor_address, 0);
> >   }
>
> this patch is causing a regression on the Clevo NL50RU of which the 
> touchpad instantly wakes up the device when going to sleep. That wasn't 
> triggered until this patch by the default settings: Setting wake capable 
> but not enabling it by default. So unless a user enabled it by hand, the 
> device went correctly to sleep.
>
> I'm not deep into this subsystem so I don't know what the best approach 
> is to ?work around this firmware bug?/?fix this issue?:
> - Changing the default back again?
> - Adding a quirk list for bad devices?
> - Maybe this isn't a firmware bug, but the touchpad was not meant to 
> wakeup the device and we can somehow detect that?

"Generally" laptops that support modern standby/s2idle can be woken from 
the touchpad.

By chance can you check Windows?  Wait a few minutes after you put it to
sleep to ensure it's really gotten down to a hardware sleep state.

> For reference: The debugging issue that lead me here: 
> https://gitlab.freedesktop.org/drm/amd/-/issues/1722#note_1719789

TLDR - That bug has multiple conflated issues, and you can ignore 95% of 
it.  It's just the last few comments we concluded and bisected down this 
regression.