Re: [PATCH v2 14/33] HID: logitech-hidpp: expose charge_type property for battery voltage feature

[Dmitry Torokhov <dmitry.torokhov@gmail.com>]

On Wed, Nov 11, 2026 at 08:39:27AM +0000, sashiko-reviews@lists.linux.dev wrote:
> [Severity: Medium]
> Does exposing POWER_SUPPLY_PROP_CHARGE_TYPE here cause sysfs to report an
> active charging rate even when the device is unplugged and discharging?
>
> hidpp20_battery_map_status_voltage() unconditionally sets *charge_type to
> POWER_SUPPLY_CHARGE_TYPE_STANDARD (or FAST / TRICKLE if bits 3 or 4 are
> set) regardless of whether external power (flags & 0x80) is present or
> status is POWER_SUPPLY_STATUS_CHARGING:
>
> drivers/hid/hid-logitech-hidpp.c:hidpp20_battery_map_status_voltage() {
>     ...
> 	} else
> 		status = POWER_SUPPLY_STATUS_DISCHARGING;
>
> 	*charge_type = POWER_SUPPLY_CHARGE_TYPE_STANDARD;
> 	if (test_bit(3, &flags)) {
> 		*charge_type = POWER_SUPPLY_CHARGE_TYPE_FAST;
> 	}
> 	if (test_bit(4, &flags)) {
> 		*charge_type = POWER_SUPPLY_CHARGE_TYPE_TRICKLE;
> 	}
>     ...
> }
>
> When a device operates wirelessly on battery power, is full, or is not
> charging, reading charge_type from sysfs returns Standard (or Fast/Trickle)
> instead of N/A (POWER_SUPPLY_CHARGE_TYPE_NONE).  Also, when a wireless
> device disconnects, hidpp_connect_event() resets online, status, and level
> to unknown, but leaves hidpp->battery.charge_type at its previous active
> value instead of POWER_SUPPLY_CHARGE_TYPE_UNKNOWN.

Agreed. I will update hidpp20_battery_map_status_voltage() in v3 so that
charge_type is only set to FAST, TRICKLE, or STANDARD when status is
POWER_SUPPLY_STATUS_CHARGING, defaults to POWER_SUPPLY_CHARGE_TYPE_NONE
when discharging, full, or not charging, and is set to
POWER_SUPPLY_CHARGE_TYPE_UNKNOWN on unknown status or wireless
disconnect.

Thanks.

-- 
Dmitry