Re: [PATCH 0/2] platform/x86: ideapad-laptop: Add charge_types:Fast (Rapid Charge)

[Jelle van der Waa]

On 11/2/25 20:24, Rong Zhang wrote:
> On Mon, 2025-11-03 at 02:57 +0800, Rong Zhang wrote:
> > Hi Jelle,
> >
> > On Sun, 2025-11-02 at 17:09 +0100, Jelle van der Waa wrote:
> > > On 10/20/25 21:24, Rong Zhang wrote:
> > > > The GBMD/SBMC interface on IdeaPad/ThinkBook supports Rapid Charge mode
> > > > (charge_types: Fast) in addition to Conservation Mode (charge_types:
> > > > Long_Life).
> > > >
> > > > This patchset exposes these two modes while carefully maintaining their
> > > > mutually exclusive state, which aligns with the behavior of manufacturer
> > > > utilities on Windows.
> > > >
> > > > Tested on ThinkBook 14 G7+ ASP.
> > >
> > > Tested this patch on my Lenovo Ideapad U330p, it now advertises that
> > > `Fast` is a supported charge_type although my laptop does not seem to
> > > support it:
> > >
> > > [root@archlinux jelle]# cat /sys/class/power_supply/BAT1/charge_types
> > > Fast [Standard] Long_Life
> > > [root@archlinux jelle]# echo 'Fast' >
> > > /sys/class/power_supply/BAT1/charge_types
> > > [root@archlinux jelle]# cat /sys/class/power_supply/BAT1/charge_types
> > > Fast [Standard] Long_Life
> >
> > Ahh, then we need an approach to determine if it is supported on a
> > specific device.
> >
> > Glancing at the disassembled DSDT.dsl of my device, I found:
> >
> >     Method (GBMD, 0, NotSerialized)
> >     {
> >     	[...]
> >     	If ((One == QCGS))
> >     	{
> >     		Local0 |= 0x00020000
> >     	}
> >     	[...]
> >     }
> >
> > BIT(17) of GBMD is 1 on my device. Maybe QCGS means "Quick CharGe
> > Supported?"
> >
> > With this assumption, I did some random Internet digging. The same bit
> > on other devices is called QKSP ("QuicK charge SuPported?"), SQCG
> > ("Support Quick CharGe?"), or QCBX (see below).
> >
> >     Method (GBMD, 0, NotSerialized)
> >     {
> >     	[...]
> >     	If ((One == QCBX))
> >     	{
> >     		If ((One == QCHO))
> >     		{
> >     			Local0 |= 0x04
> >     		}
> >     	}
> >     	[...]
> >     	If ((One == QCBX))
> >     	{
> >     		Local0 |= 0x00020000
> >     	}
> >     	[...]
> >     }
> >
> > https://badland.io/static/acpidump.txt
> >
> > 0x04 is BIT(2)/GBMD_RAPID_CHARGE_STATE_BIT. With all these pieces of
> > information, I presume BIT(17) of GBMD is what we are searching for.
> >
> > > I'm wondering if the battery extension API allows to not advertise a
> > > property if it isn't supported or if it should at least return -EINVAL.
> >
> > We can achieve this by defining multiple struct power_supply_ext. See
> > drivers/power/supply/cros_charge-control.c.
> >
> > Could you test the patch below (based on "review-ilpo-next")?
>
> Note: this patch is just a quick PoC (I am going to sleep now, zzz...).
> ideapad_psy_ext_{get,set}_prop need to be reorganized to properly
> support your device. If `cat charge_types' doesn't show `Fast', we're
> in the right direction.

Thanks for the quick patch!

Tested it on my Ideapad U330p:

[root@archlinux ~]# cat /sys/class/power_supply/BAT1/charge_types
[Standard] Long_Life
[root@archlinux ~]# echo Long_Life > 
/sys/class/power_supply/BAT1/charge_types
[root@archlinux ~]# cat /sys/class/power_supply/BAT1/charge_types
Standard [Long_Life]

I'm however still waiting on the laptop to slowly charge to 80% to 
confirm that charge limits are applied.

Greetings,

Jelle van der Waa