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.

Rong Zhang (2):
  platform/x86: ideapad-laptop: Protect GBMD/SBMC calls with mutex
  platform/x86: ideapad-laptop: Add charge_types [Fast] (Rapid Charge)

 drivers/platform/x86/lenovo/ideapad-laptop.c | 144 ++++++++++++++-----
 1 file changed, 108 insertions(+), 36 deletions(-)


base-commit: 211ddde0823f1442e4ad052a2f30f050145ccada
-- 
2.51.0

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

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.

Greetings,

Jelle van der Waa

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")?

@Ilpo:

This patch series has been merge into your "review-ilpo-next" branch.

Should I reorganize the series and send a [PATCH v2]? Or should I just
send the patch below (after adding a commit message, ofc)?

> Greetings,
> 
> Jelle van der Waa

Thanks,
Rong

---
diff --git a/drivers/platform/x86/lenovo/ideapad-laptop.c b/drivers/platform/x86/lenovo/ideapad-laptop.c
index 931a72a2a487..b9927493cb93 100644
--- a/drivers/platform/x86/lenovo/ideapad-laptop.c
+++ b/drivers/platform/x86/lenovo/ideapad-laptop.c
@@ -75,6 +75,7 @@ enum {
 enum {
 	GBMD_RAPID_CHARGE_STATE_BIT = 2,
 	GBMD_CONSERVATION_STATE_BIT = 5,
+	GBMD_RAPID_CHARGE_SUPPORTED_BIT = 17,
 };
 
 enum {
@@ -180,6 +181,7 @@ struct ideapad_private {
 	struct ideapad_dytc_priv *dytc;
 	struct dentry *debug;
 	struct acpi_battery_hook battery_hook;
+	const struct power_supply_ext *battery_ext;
 	unsigned long cfg;
 	unsigned long r_touchpad_val;
 	struct {
@@ -2119,30 +2121,42 @@ static const enum power_supply_property ideapad_power_supply_props[] = {
 	POWER_SUPPLY_PROP_CHARGE_TYPES,
 };
 
-static const struct power_supply_ext ideapad_battery_ext = {
-	.name			= "ideapad_laptop",
-	.properties		= ideapad_power_supply_props,
-	.num_properties		= ARRAY_SIZE(ideapad_power_supply_props),
-	.charge_types		= (BIT(POWER_SUPPLY_CHARGE_TYPE_STANDARD) |
-				   BIT(POWER_SUPPLY_CHARGE_TYPE_FAST) |
-				   BIT(POWER_SUPPLY_CHARGE_TYPE_LONGLIFE)),
-	.get_property		= ideapad_psy_ext_get_prop,
-	.set_property		= ideapad_psy_ext_set_prop,
-	.property_is_writeable	= ideapad_psy_prop_is_writeable,
-};
+#define DEFINE_IDEAPAD_POWER_SUPPLY_EXTENSION(_name, _charge_types)			\
+	static const struct power_supply_ext _name = {					\
+		.name			= "ideapad_laptop",				\
+		.properties		= ideapad_power_supply_props,			\
+		.num_properties		= ARRAY_SIZE(ideapad_power_supply_props),	\
+		.charge_types		= _charge_types,				\
+		.get_property		= ideapad_psy_ext_get_prop,			\
+		.set_property		= ideapad_psy_ext_set_prop,			\
+		.property_is_writeable	= ideapad_psy_prop_is_writeable,		\
+	}
+
+DEFINE_IDEAPAD_POWER_SUPPLY_EXTENSION(ideapad_battery_ext_v1,
+	(BIT(POWER_SUPPLY_CHARGE_TYPE_STANDARD) |
+	 BIT(POWER_SUPPLY_CHARGE_TYPE_LONGLIFE))
+);
+
+DEFINE_IDEAPAD_POWER_SUPPLY_EXTENSION(ideapad_battery_ext_v2,
+	(BIT(POWER_SUPPLY_CHARGE_TYPE_STANDARD) |
+	 BIT(POWER_SUPPLY_CHARGE_TYPE_FAST) |
+	 BIT(POWER_SUPPLY_CHARGE_TYPE_LONGLIFE))
+);
 
 static int ideapad_battery_add(struct power_supply *battery, struct acpi_battery_hook *hook)
 {
 	struct ideapad_private *priv = container_of(hook, struct ideapad_private, battery_hook);
 
-	return power_supply_register_extension(battery, &ideapad_battery_ext,
+	return power_supply_register_extension(battery, priv->battery_ext,
 					       &priv->platform_device->dev, priv);
 }
 
 static int ideapad_battery_remove(struct power_supply *battery,
 				  struct acpi_battery_hook *hook)
 {
-	power_supply_unregister_extension(battery, &ideapad_battery_ext);
+	struct ideapad_private *priv = container_of(hook, struct ideapad_private, battery_hook);
+
+	power_supply_unregister_extension(battery, priv->battery_ext);
 
 	return 0;
 }
@@ -2167,14 +2181,22 @@ static int ideapad_check_features(struct ideapad_private *priv)
 		priv->features.fan_mode = true;
 
 	if (acpi_has_method(handle, "GBMD") && acpi_has_method(handle, "SBMC")) {
-		priv->features.conservation_mode = true;
-		priv->battery_hook.add_battery = ideapad_battery_add;
-		priv->battery_hook.remove_battery = ideapad_battery_remove;
-		priv->battery_hook.name = "Ideapad Battery Extension";
-
-		err = devm_battery_hook_register(&priv->platform_device->dev, &priv->battery_hook);
-		if (err)
-			return err;
+		/* Not acquiring gbmd_sbmc_mutex as race condition is impossible on init */
+		if (!eval_gbmd(handle, &val)) {
+			priv->features.conservation_mode = true;
+			priv->battery_ext = test_bit(GBMD_RAPID_CHARGE_SUPPORTED_BIT, &val)
+					  ? &ideapad_battery_ext_v2
+					  : &ideapad_battery_ext_v1;
+
+			priv->battery_hook.add_battery = ideapad_battery_add;
+			priv->battery_hook.remove_battery = ideapad_battery_remove;
+			priv->battery_hook.name = "Ideapad Battery Extension";
+
+			err = devm_battery_hook_register(&priv->platform_device->dev,
+							 &priv->battery_hook);
+			if (err)
+				return err;
+		}
 	}
 
 	if (acpi_has_method(handle, "DYTC"))

On Mon, 3 Nov 2025, 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")?
> 
> @Ilpo:
> 
> This patch series has been merge into your "review-ilpo-next" branch.
> 
> Should I reorganize the series and send a [PATCH v2]? Or should I just
> send the patch below (after adding a commit message, ofc)?

Hi Rong,

I've dropped those two v1 patches from the review-ilpo-next branch.

(They are still in for-next but will be gone from there as well on the 
next review-ilpo-next -> for-next propagation.)

So please send the full series with this fixed as v2. You can include the 
str_on_off() change also within the v2 series.

-- 
 i.

> > Greetings,
> > 
> > Jelle van der Waa
> 
> Thanks,
> Rong
> 
> ---
> diff --git a/drivers/platform/x86/lenovo/ideapad-laptop.c b/drivers/platform/x86/lenovo/ideapad-laptop.c
> index 931a72a2a487..b9927493cb93 100644
> --- a/drivers/platform/x86/lenovo/ideapad-laptop.c
> +++ b/drivers/platform/x86/lenovo/ideapad-laptop.c
> @@ -75,6 +75,7 @@ enum {
>  enum {
>  	GBMD_RAPID_CHARGE_STATE_BIT = 2,
>  	GBMD_CONSERVATION_STATE_BIT = 5,
> +	GBMD_RAPID_CHARGE_SUPPORTED_BIT = 17,
>  };
>  
>  enum {
> @@ -180,6 +181,7 @@ struct ideapad_private {
>  	struct ideapad_dytc_priv *dytc;
>  	struct dentry *debug;
>  	struct acpi_battery_hook battery_hook;
> +	const struct power_supply_ext *battery_ext;
>  	unsigned long cfg;
>  	unsigned long r_touchpad_val;
>  	struct {
> @@ -2119,30 +2121,42 @@ static const enum power_supply_property ideapad_power_supply_props[] = {
>  	POWER_SUPPLY_PROP_CHARGE_TYPES,
>  };
>  
> -static const struct power_supply_ext ideapad_battery_ext = {
> -	.name			= "ideapad_laptop",
> -	.properties		= ideapad_power_supply_props,
> -	.num_properties		= ARRAY_SIZE(ideapad_power_supply_props),
> -	.charge_types		= (BIT(POWER_SUPPLY_CHARGE_TYPE_STANDARD) |
> -				   BIT(POWER_SUPPLY_CHARGE_TYPE_FAST) |
> -				   BIT(POWER_SUPPLY_CHARGE_TYPE_LONGLIFE)),
> -	.get_property		= ideapad_psy_ext_get_prop,
> -	.set_property		= ideapad_psy_ext_set_prop,
> -	.property_is_writeable	= ideapad_psy_prop_is_writeable,
> -};
> +#define DEFINE_IDEAPAD_POWER_SUPPLY_EXTENSION(_name, _charge_types)			\
> +	static const struct power_supply_ext _name = {					\
> +		.name			= "ideapad_laptop",				\
> +		.properties		= ideapad_power_supply_props,			\
> +		.num_properties		= ARRAY_SIZE(ideapad_power_supply_props),	\
> +		.charge_types		= _charge_types,				\
> +		.get_property		= ideapad_psy_ext_get_prop,			\
> +		.set_property		= ideapad_psy_ext_set_prop,			\
> +		.property_is_writeable	= ideapad_psy_prop_is_writeable,		\
> +	}
> +
> +DEFINE_IDEAPAD_POWER_SUPPLY_EXTENSION(ideapad_battery_ext_v1,
> +	(BIT(POWER_SUPPLY_CHARGE_TYPE_STANDARD) |
> +	 BIT(POWER_SUPPLY_CHARGE_TYPE_LONGLIFE))
> +);
> +
> +DEFINE_IDEAPAD_POWER_SUPPLY_EXTENSION(ideapad_battery_ext_v2,
> +	(BIT(POWER_SUPPLY_CHARGE_TYPE_STANDARD) |
> +	 BIT(POWER_SUPPLY_CHARGE_TYPE_FAST) |
> +	 BIT(POWER_SUPPLY_CHARGE_TYPE_LONGLIFE))
> +);
>  
>  static int ideapad_battery_add(struct power_supply *battery, struct acpi_battery_hook *hook)
>  {
>  	struct ideapad_private *priv = container_of(hook, struct ideapad_private, battery_hook);
>  
> -	return power_supply_register_extension(battery, &ideapad_battery_ext,
> +	return power_supply_register_extension(battery, priv->battery_ext,
>  					       &priv->platform_device->dev, priv);
>  }
>  
>  static int ideapad_battery_remove(struct power_supply *battery,
>  				  struct acpi_battery_hook *hook)
>  {
> -	power_supply_unregister_extension(battery, &ideapad_battery_ext);
> +	struct ideapad_private *priv = container_of(hook, struct ideapad_private, battery_hook);
> +
> +	power_supply_unregister_extension(battery, priv->battery_ext);
>  
>  	return 0;
>  }
> @@ -2167,14 +2181,22 @@ static int ideapad_check_features(struct ideapad_private *priv)
>  		priv->features.fan_mode = true;
>  
>  	if (acpi_has_method(handle, "GBMD") && acpi_has_method(handle, "SBMC")) {
> -		priv->features.conservation_mode = true;
> -		priv->battery_hook.add_battery = ideapad_battery_add;
> -		priv->battery_hook.remove_battery = ideapad_battery_remove;
> -		priv->battery_hook.name = "Ideapad Battery Extension";
> -
> -		err = devm_battery_hook_register(&priv->platform_device->dev, &priv->battery_hook);
> -		if (err)
> -			return err;
> +		/* Not acquiring gbmd_sbmc_mutex as race condition is impossible on init */
> +		if (!eval_gbmd(handle, &val)) {
> +			priv->features.conservation_mode = true;
> +			priv->battery_ext = test_bit(GBMD_RAPID_CHARGE_SUPPORTED_BIT, &val)
> +					  ? &ideapad_battery_ext_v2
> +					  : &ideapad_battery_ext_v1;
> +
> +			priv->battery_hook.add_battery = ideapad_battery_add;
> +			priv->battery_hook.remove_battery = ideapad_battery_remove;
> +			priv->battery_hook.name = "Ideapad Battery Extension";
> +
> +			err = devm_battery_hook_register(&priv->platform_device->dev,
> +							 &priv->battery_hook);
> +			if (err)
> +				return err;
> +		}
>  	}
>  
>  	if (acpi_has_method(handle, "DYTC"))
>

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,
Rong

> @Ilpo:
> 
> This patch series has been merge into your "review-ilpo-next" branch.
> 
> Should I reorganize the series and send a [PATCH v2]? Or should I just
> send the patch below (after adding a commit message, ofc)?
> 
> > Greetings,
> > 
> > Jelle van der Waa
> 
> Thanks,
> Rong
> 
> ---
> diff --git a/drivers/platform/x86/lenovo/ideapad-laptop.c b/drivers/platform/x86/lenovo/ideapad-laptop.c
> index 931a72a2a487..b9927493cb93 100644
> --- a/drivers/platform/x86/lenovo/ideapad-laptop.c
> +++ b/drivers/platform/x86/lenovo/ideapad-laptop.c
> @@ -75,6 +75,7 @@ enum {
>  enum {
>  	GBMD_RAPID_CHARGE_STATE_BIT = 2,
>  	GBMD_CONSERVATION_STATE_BIT = 5,
> +	GBMD_RAPID_CHARGE_SUPPORTED_BIT = 17,
>  };
>  
>  enum {
> @@ -180,6 +181,7 @@ struct ideapad_private {
>  	struct ideapad_dytc_priv *dytc;
>  	struct dentry *debug;
>  	struct acpi_battery_hook battery_hook;
> +	const struct power_supply_ext *battery_ext;
>  	unsigned long cfg;
>  	unsigned long r_touchpad_val;
>  	struct {
> @@ -2119,30 +2121,42 @@ static const enum power_supply_property ideapad_power_supply_props[] = {
>  	POWER_SUPPLY_PROP_CHARGE_TYPES,
>  };
>  
> -static const struct power_supply_ext ideapad_battery_ext = {
> -	.name			= "ideapad_laptop",
> -	.properties		= ideapad_power_supply_props,
> -	.num_properties		= ARRAY_SIZE(ideapad_power_supply_props),
> -	.charge_types		= (BIT(POWER_SUPPLY_CHARGE_TYPE_STANDARD) |
> -				   BIT(POWER_SUPPLY_CHARGE_TYPE_FAST) |
> -				   BIT(POWER_SUPPLY_CHARGE_TYPE_LONGLIFE)),
> -	.get_property		= ideapad_psy_ext_get_prop,
> -	.set_property		= ideapad_psy_ext_set_prop,
> -	.property_is_writeable	= ideapad_psy_prop_is_writeable,
> -};
> +#define DEFINE_IDEAPAD_POWER_SUPPLY_EXTENSION(_name, _charge_types)			\
> +	static const struct power_supply_ext _name = {					\
> +		.name			= "ideapad_laptop",				\
> +		.properties		= ideapad_power_supply_props,			\
> +		.num_properties		= ARRAY_SIZE(ideapad_power_supply_props),	\
> +		.charge_types		= _charge_types,				\
> +		.get_property		= ideapad_psy_ext_get_prop,			\
> +		.set_property		= ideapad_psy_ext_set_prop,			\
> +		.property_is_writeable	= ideapad_psy_prop_is_writeable,		\
> +	}
> +
> +DEFINE_IDEAPAD_POWER_SUPPLY_EXTENSION(ideapad_battery_ext_v1,
> +	(BIT(POWER_SUPPLY_CHARGE_TYPE_STANDARD) |
> +	 BIT(POWER_SUPPLY_CHARGE_TYPE_LONGLIFE))
> +);
> +
> +DEFINE_IDEAPAD_POWER_SUPPLY_EXTENSION(ideapad_battery_ext_v2,
> +	(BIT(POWER_SUPPLY_CHARGE_TYPE_STANDARD) |
> +	 BIT(POWER_SUPPLY_CHARGE_TYPE_FAST) |
> +	 BIT(POWER_SUPPLY_CHARGE_TYPE_LONGLIFE))
> +);
>  
>  static int ideapad_battery_add(struct power_supply *battery, struct acpi_battery_hook *hook)
>  {
>  	struct ideapad_private *priv = container_of(hook, struct ideapad_private, battery_hook);
>  
> -	return power_supply_register_extension(battery, &ideapad_battery_ext,
> +	return power_supply_register_extension(battery, priv->battery_ext,
>  					       &priv->platform_device->dev, priv);
>  }
>  
>  static int ideapad_battery_remove(struct power_supply *battery,
>  				  struct acpi_battery_hook *hook)
>  {
> -	power_supply_unregister_extension(battery, &ideapad_battery_ext);
> +	struct ideapad_private *priv = container_of(hook, struct ideapad_private, battery_hook);
> +
> +	power_supply_unregister_extension(battery, priv->battery_ext);
>  
>  	return 0;
>  }
> @@ -2167,14 +2181,22 @@ static int ideapad_check_features(struct ideapad_private *priv)
>  		priv->features.fan_mode = true;
>  
>  	if (acpi_has_method(handle, "GBMD") && acpi_has_method(handle, "SBMC")) {
> -		priv->features.conservation_mode = true;
> -		priv->battery_hook.add_battery = ideapad_battery_add;
> -		priv->battery_hook.remove_battery = ideapad_battery_remove;
> -		priv->battery_hook.name = "Ideapad Battery Extension";
> -
> -		err = devm_battery_hook_register(&priv->platform_device->dev, &priv->battery_hook);
> -		if (err)
> -			return err;
> +		/* Not acquiring gbmd_sbmc_mutex as race condition is impossible on init */
> +		if (!eval_gbmd(handle, &val)) {
> +			priv->features.conservation_mode = true;
> +			priv->battery_ext = test_bit(GBMD_RAPID_CHARGE_SUPPORTED_BIT, &val)
> +					  ? &ideapad_battery_ext_v2
> +					  : &ideapad_battery_ext_v1;
> +
> +			priv->battery_hook.add_battery = ideapad_battery_add;
> +			priv->battery_hook.remove_battery = ideapad_battery_remove;
> +			priv->battery_hook.name = "Ideapad Battery Extension";
> +
> +			err = devm_battery_hook_register(&priv->platform_device->dev,
> +							 &priv->battery_hook);
> +			if (err)
> +				return err;
> +		}
>  	}
>  
>  	if (acpi_has_method(handle, "DYTC"))

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