Skip to contentSkip to conversation

HID: logitech-hidpp: add support for Unified Battery (1004) feature

PATCH In review [email protected] 3 participants 9 messages Jan 4, 2021 – Jan 18, 2021

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 HID: logitech-hidpp: add support for Unified Battery (1004) feature

From: Filipe Laíns <[email protected]>

This new feature present in new devices replaces the old Battery Level Status (0x1000) feature. It keeps essentially the same information for levels (reporting critical, low, good and full) but makes these levels optional, the device exports a capability setting which describes which levels it supports. In addition to this, there is an optional state_of_charge paramenter that exports the battery percentage.

This patch adds support for this new feature. There were some implementation choices, as described below and in the code.

If the device supports the state_of_charge parameter, we will just export the battery percentage and not the levels, which the device might still support.

Since this feature can co-exist with the Battery Voltage (0x1001) feature and we currently only support one battery feature, I changed the battery feature discovery to try to use 0x1000 and 0x1004 first and only then 0x1001, the battery voltage feature.
In the future we could uncouple this and make the battery feature co-exists with 0x1000 and 0x1004, allowing the device to export voltage information in addition to the battery percentage or level.

I tested this patch with a MX Anywhere 3, which supports the new feature. Since I don't have any device that doesn't support the state_of_charge parameter of this feature, I forced the MX Anywhere 3 to use the level information, instead of battery percentage, to test that part of the implementation.
I also tested with a MX Master 3, which supports the Battery Level Status (0x1000) feature, and a G703 Hero, which supports the Battery Voltage (0x1001) feature, to make sure nothing broke there.

---
 drivers/hid/hid-logitech-hidpp.c | 244 ++++++++++++++++++++++++++++++-
 1 file changed, 237 insertions(+), 7 deletions(-)
drivers/hid/hid-logitech-hidpp.c +237 -7
index f85781464807..291c6b4d26b7 100644
@@ -92,6 +92,8 @@ MODULE_PARM_DESC(disable_tap_to_click,
 #define HIDPP_CAPABILITY_BATTERY_MILEAGE	BIT(2)
 #define HIDPP_CAPABILITY_BATTERY_LEVEL_STATUS	BIT(3)
 #define HIDPP_CAPABILITY_BATTERY_VOLTAGE	BIT(4)
+#define HIDPP_CAPABILITY_BATTERY_PERCENTAGE	BIT(5)
+#define HIDPP_CAPABILITY_UNIFIED_BATTERY	BIT(6)
 
 #define lg_map_key_clear(c)  hid_map_usage_clear(hi, usage, bit, max, EV_KEY, (c))
 
@@ -152,6 +154,7 @@ struct hidpp_battery {
 	int voltage;
 	int charge_type;
 	bool online;
+	u8 supported_levels_1004;
 };
 
 /**
@@ -1171,7 +1174,7 @@ static int hidpp20_batterylevel_get_battery_info(struct hidpp_device *hidpp,
 	return 0;
 }
 
-static int hidpp20_query_battery_info(struct hidpp_device *hidpp)
+static int hidpp20_query_battery_info_1000(struct hidpp_device *hidpp)
 {
 	u8 feature_type;
 	int ret;
@@ -1208,7 +1211,7 @@ static int hidpp20_query_battery_info(struct hidpp_device *hidpp)
 	return 0;
 }
 
-static int hidpp20_battery_event(struct hidpp_device *hidpp,
+static int hidpp20_battery_event_1000(struct hidpp_device *hidpp,
 				 u8 *data, int size)
 {
 	struct hidpp_report *report = (struct hidpp_report *)data;
@@ -1380,6 +1383,222 @@ static int hidpp20_battery_voltage_event(struct hidpp_device *hidpp,
 	return 0;
 }
 
+/* -------------------------------------------------------------------------- */
+/* 0x1004: Unified battery                                                    */
+/* -------------------------------------------------------------------------- */
+
+#define HIDPP_PAGE_UNIFIED_BATTERY				0x1004
+
+#define CMD_UNIFIED_BATTERY_GET_CAPABILITIES			0x00
+#define CMD_UNIFIED_BATTERY_GET_STATUS				0x10
+
+#define EVENT_UNIFIED_BATTERY_STATUS_EVENT			0x00
+
+#define FLAG_UNIFIED_BATTERY_LEVEL_CRITICAL			BIT(0)
+#define FLAG_UNIFIED_BATTERY_LEVEL_LOW				BIT(1)
+#define FLAG_UNIFIED_BATTERY_LEVEL_GOOD				BIT(2)
+#define FLAG_UNIFIED_BATTERY_LEVEL_FULL				BIT(3)
+
+#define FLAG_UNIFIED_BATTERY_FLAGS_RECHARGEABLE			BIT(0)
+#define FLAG_UNIFIED_BATTERY_FLAGS_STATE_OF_CHARGE		BIT(1)
+
+static int hidpp20_unifiedbattery_get_capabilities(struct hidpp_device *hidpp,
+						   u8 feature_index)
+{
+	struct hidpp_report response;
+	int ret;
+	u8 *params = (u8 *)response.fap.params;
+
+	if (hidpp->capabilities & HIDPP_CAPABILITY_BATTERY_LEVEL_STATUS ||
+	    hidpp->capabilities & HIDPP_CAPABILITY_BATTERY_PERCENTAGE) {
+		/* we have already set the device capabilities, so let's skip */
+		return 0;
+	}
+
+	ret = hidpp_send_fap_command_sync(hidpp, feature_index,
+					  CMD_UNIFIED_BATTERY_GET_CAPABILITIES,
+					  NULL, 0, &response);
+	/* Ignore these intermittent errors */
+	if (ret == HIDPP_ERROR_RESOURCE_ERROR)
+		return -EIO;
+	if (ret > 0) {
+		hid_err(hidpp->hid_dev, "%s: received protocol error 0x%02x\n",
+			__func__, ret);
+		return -EPROTO;
+	}
+	if (ret)
+		return ret;
+
+	/* if the device supports state of charge (battery percentage) we won't
+	   export the battery level information. there are 4 possible battery
+	   levels and they all are optional, this means that the device might
+	   not support any of them, we are just better off with the battery
+	   percentage. */
+	if (params[1] & FLAG_UNIFIED_BATTERY_FLAGS_STATE_OF_CHARGE) {
+		hidpp->capabilities |= HIDPP_CAPABILITY_BATTERY_PERCENTAGE;
+		hidpp->battery.supported_levels_1004 = 0;
+	} else {
+		hidpp->capabilities |= HIDPP_CAPABILITY_BATTERY_LEVEL_STATUS;
+		hidpp->battery.supported_levels_1004 = params[0];
+	}
+
+	return 0;
+}
+
+static int hidpp20_unifiedbattery_map_status(struct hidpp_device *hidpp,
+					     u8 charging_status,
+					     u8 external_power_status)
+{
+	int status;
+
+	switch (charging_status) {
+		case 0: /* discharging */
+			status = POWER_SUPPLY_STATUS_DISCHARGING;
+			break;
+		case 1: /* charging */
+		case 2: /* charging slow */
+			status = POWER_SUPPLY_STATUS_CHARGING;
+			break;
+		case 3: /* complete */
+			status = POWER_SUPPLY_STATUS_FULL;
+			break;
+		case 4: /* error */
+			status = POWER_SUPPLY_STATUS_NOT_CHARGING;
+			hid_info(hidpp->hid_dev, "%s: charging error",
+				 hidpp->name);
+			break;
+		default:
+			status = POWER_SUPPLY_STATUS_NOT_CHARGING;
+			break;
+	}
+
+	return status;
+}
+
+static int hidpp20_unifiedbattery_map_level(struct hidpp_device *hidpp,
+					    u8 battery_level)
+{
+	/* cler unsupported level bits */
+	battery_level &= hidpp->battery.supported_levels_1004;
+
+	if (battery_level & FLAG_UNIFIED_BATTERY_LEVEL_FULL)
+		return POWER_SUPPLY_CAPACITY_LEVEL_FULL;
+	else if (battery_level & FLAG_UNIFIED_BATTERY_LEVEL_GOOD)
+		return POWER_SUPPLY_CAPACITY_LEVEL_NORMAL;
+	else if (battery_level & FLAG_UNIFIED_BATTERY_LEVEL_LOW)
+		return POWER_SUPPLY_CAPACITY_LEVEL_LOW;
+	else if (battery_level & FLAG_UNIFIED_BATTERY_LEVEL_CRITICAL)
+		return POWER_SUPPLY_CAPACITY_LEVEL_CRITICAL;
+
+	return POWER_SUPPLY_CAPACITY_LEVEL_UNKNOWN;
+}
+
+static int hidpp20_unifiedbattery_get_status(struct hidpp_device *hidpp,
+					     u8 feature_index,
+					     u8 *state_of_charge,
+					     int *status,
+					     int *level)
+{
+	struct hidpp_report response;
+	int ret;
+	u8 *params = (u8 *)response.fap.params;
+
+	ret = hidpp_send_fap_command_sync(hidpp, feature_index,
+					  CMD_UNIFIED_BATTERY_GET_STATUS,
+					  NULL, 0, &response);
+	/* Ignore these intermittent errors */
+	if (ret == HIDPP_ERROR_RESOURCE_ERROR)
+		return -EIO;
+	if (ret > 0) {
+		hid_err(hidpp->hid_dev, "%s: received protocol error 0x%02x\n",
+			__func__, ret);
+		return -EPROTO;
+	}
+	if (ret)
+		return ret;
+
+	*state_of_charge = params[0];
+	*status = hidpp20_unifiedbattery_map_status(hidpp, params[2], params[3]);
+	*level = hidpp20_unifiedbattery_map_level(hidpp, params[1]);
+
+	return 0;
+}
+
+static int hidpp20_query_battery_info_1004(struct hidpp_device *hidpp)
+{
+	u8 feature_type;
+	int ret;
+	u8 state_of_charge;
+	int status, level;
+
+	if (hidpp->battery.feature_index == 0xff) {
+		ret = hidpp_root_get_feature(hidpp,
+					     HIDPP_PAGE_UNIFIED_BATTERY,
+					     &hidpp->battery.feature_index,
+					     &feature_type);
+		if (ret)
+			return ret;
+	}
+
+	ret = hidpp20_unifiedbattery_get_capabilities(hidpp,
+					hidpp->battery.feature_index);
+	if (ret)
+		return ret;
+
+	ret = hidpp20_unifiedbattery_get_status(hidpp,
+						hidpp->battery.feature_index,
+						&state_of_charge,
+						&status,
+						&level);
+	if (ret)
+		return ret;
+
+	hidpp->capabilities |= HIDPP_CAPABILITY_UNIFIED_BATTERY;
+	hidpp->battery.capacity = state_of_charge;
+	hidpp->battery.status = status;
+	hidpp->battery.level = level;
+	hidpp->battery.online = true;
+
+	return 0;
+}
+
+static int hidpp20_battery_event_1004(struct hidpp_device *hidpp,
+				 u8 *data, int size)
+{
+	struct hidpp_report *report = (struct hidpp_report *)data;
+	u8 *params = (u8 *)report->fap.params;
+	int state_of_charge, status, level;
+	bool changed;
+
+	if (report->fap.feature_index != hidpp->battery.feature_index ||
+	    report->fap.funcindex_clientid != EVENT_UNIFIED_BATTERY_STATUS_EVENT)
+		return 0;
+
+	state_of_charge = params[0];
+	status = hidpp20_unifiedbattery_map_status(hidpp, params[2], params[3]);
+	level = hidpp20_unifiedbattery_map_level(hidpp, params[1]);
+
+	changed = status != hidpp->battery.status ||
+		  (state_of_charge != hidpp->battery.capacity &&
+		   hidpp->capabilities & HIDPP_CAPABILITY_BATTERY_PERCENTAGE) ||
+		  (level != hidpp->battery.level &&
+		   hidpp->capabilities & HIDPP_CAPABILITY_BATTERY_LEVEL_STATUS);
+
+	if (changed) {
+		hidpp->battery.capacity = state_of_charge;
+		hidpp->battery.status = status;
+		hidpp->battery.level = level;
+		if (hidpp->battery.ps)
+			power_supply_changed(hidpp->battery.ps);
+	}
+
+	return 0;
+}
+
+/* -------------------------------------------------------------------------- */
+/* Battery feature helpers                                                    */
+/* -------------------------------------------------------------------------- */
+
 static enum power_supply_property hidpp_battery_props[] = {
 	POWER_SUPPLY_PROP_ONLINE,
 	POWER_SUPPLY_PROP_STATUS,
@@ -3307,7 +3526,10 @@ static int hidpp_raw_hidpp_event(struct hidpp_device *hidpp, u8 *data,
 	}
 
 	if (hidpp->capabilities & HIDPP_CAPABILITY_HIDPP20_BATTERY) {
-		ret = hidpp20_battery_event(hidpp, data, size);
+		ret = hidpp20_battery_event_1000(hidpp, data, size);
+		if (ret != 0)
+			return ret;
+		ret = hidpp20_battery_event_1004(hidpp, data, size);
 		if (ret != 0)
 			return ret;
 		ret = hidpp_solar_battery_event(hidpp, data, size);
@@ -3443,9 +3665,14 @@ static int hidpp_initialize_battery(struct hidpp_device *hidpp)
 		if (hidpp->quirks & HIDPP_QUIRK_CLASS_K750)
 			ret = hidpp_solar_request_battery_event(hidpp);
 		else {
-			ret = hidpp20_query_battery_voltage_info(hidpp);
+			/* we only support one battery feature right now, so let's
+			   first check the ones that support battery level first
+			   and leave voltage for last */
+			ret = hidpp20_query_battery_info_1000(hidpp);
+			if (ret)
+				ret = hidpp20_query_battery_info_1004(hidpp);
 			if (ret)
-				ret = hidpp20_query_battery_info(hidpp);
+				ret = hidpp20_query_battery_voltage_info(hidpp);
 		}
 
 		if (ret)
@@ -3473,7 +3700,8 @@ static int hidpp_initialize_battery(struct hidpp_device *hidpp)
 
 	num_battery_props = ARRAY_SIZE(hidpp_battery_props) - 3;
 
-	if (hidpp->capabilities & HIDPP_CAPABILITY_BATTERY_MILEAGE)
+	if (hidpp->capabilities & HIDPP_CAPABILITY_BATTERY_MILEAGE ||
+	    hidpp->capabilities & HIDPP_CAPABILITY_BATTERY_PERCENTAGE)
 		battery_props[num_battery_props++] =
 				POWER_SUPPLY_PROP_CAPACITY;
 
@@ -3650,8 +3878,10 @@ static void hidpp_connect_event(struct hidpp_device *hidpp)
 	} else if (hidpp->capabilities & HIDPP_CAPABILITY_HIDPP20_BATTERY) {
 		if (hidpp->capabilities & HIDPP_CAPABILITY_BATTERY_VOLTAGE)
 			hidpp20_query_battery_voltage_info(hidpp);
+		else if (hidpp->capabilities & HIDPP_CAPABILITY_UNIFIED_BATTERY)
+			hidpp20_query_battery_info_1004(hidpp);
 		else
-			hidpp20_query_battery_info(hidpp);
+			hidpp20_query_battery_info_1000(hidpp);
 	}
 	if (hidpp->battery.ps)
 		power_supply_changed(hidpp->battery.ps);
Signature · 1 line
2.30.0
#1 lore raw
Quoted 13 lines · On Mon, 2021-01-04 at 18:29 +0000, [email protected] wrote:

From: Filipe Laíns <[email protected]>

This new feature present in new devices replaces the old Battery Level
Status (0x1000) feature. It keeps essentially the same information for
levels (reporting critical, low, good and full) but makes these levels
optional, the device exports a capability setting which describes which
levels it supports. In addition to this, there is an optional state_of_charge paramenter that exports the battery percentage.

Quoted 7 lines

This patch adds support for this new feature. There were some implementation choices, as described below and in the code.

If the device supports the state_of_charge parameter, we will just export the battery percentage and not the levels, which the device might
still support.

I'm guessing that means no changes needed on the upower side?

Cheers

#2 lore raw
Quoted 17 lines · Bastien Nocera
Quoted 16 lines · On Mon, 2021-01-04 at 18:29 +0000, [email protected] wrote:

From: Filipe Laíns <[email protected]>

This new feature present in new devices replaces the old Battery Level
Status (0x1000) feature. It keeps essentially the same information for
levels (reporting critical, low, good and full) but makes these levels
optional, the device exports a capability setting which describes which
levels it supports. In addition to this, there is an optional state_of_charge paramenter that exports the battery percentage.

This patch adds support for this new feature. There were some implementation choices, as described below and in the code.

Quoted 9 lines
Quoted 4 lines

If the device supports the state_of_charge parameter, we will just export the battery percentage and not the levels, which the device might
still support.

I'm guessing that means no changes needed on the upower side?

Cheers

Yes :)
I tested upower and all works as expected.

There will still be devices that only support battery voltage, so I might implement the battery voltage to charge percentage in a future patch.

Cheers,
Filipe Laíns

#3 lore raw
Quoted 31 lines · Filipe Laíns
Quoted 25 lines · Bastien Nocera
Quoted 17 lines · On Mon, 2021-01-04 at 18:29 +0000, [email protected] wrote:

From: Filipe Laíns <[email protected]>

This new feature present in new devices replaces the old Battery Level
Status (0x1000) feature. It keeps essentially the same information
for
levels (reporting critical, low, good and full) but makes these levels
optional, the device exports a capability setting which describes which
levels it supports. In addition to this, there is an optional state_of_charge paramenter that exports the battery percentage.

This patch adds support for this new feature. There were some implementation choices, as described below and in the code.

Quoted 6 lines

If the device supports the state_of_charge parameter, we will just
export the battery percentage and not the levels, which the device
might
still support.

Quoted 4 lines

I'm guessing that means no changes needed on the upower side?

Cheers

Quoted 6 lines

Yes :)
I tested upower and all works as expected.

There will still be devices that only support battery voltage, so I might
implement the battery voltage to charge percentage in a future patch.

I sent a WIP patch at the end of November for that, it wasn't even compile-tested, but might be a good base to start from.

I don't think that I have any hardware that supports that feature in any case.

Cheers

maintainer
#4 lore raw
In reply to [email protected]
Quoted 24 lines · On Mon, 4 Jan 2021, [email protected] wrote:

From: Filipe Laíns <[email protected]>

This new feature present in new devices replaces the old Battery Level Status (0x1000) feature. It keeps essentially the same information for levels (reporting critical, low, good and full) but makes these levels optional, the device exports a capability setting which describes which levels it supports. In addition to this, there is an optional state_of_charge paramenter that exports the battery percentage.

This patch adds support for this new feature. There were some implementation choices, as described below and in the code.

If the device supports the state_of_charge parameter, we will just export the battery percentage and not the levels, which the device might still support.

Since this feature can co-exist with the Battery Voltage (0x1001) feature and we currently only support one battery feature, I changed the battery feature discovery to try to use 0x1000 and 0x1004 first and only then 0x1001, the battery voltage feature.
In the future we could uncouple this and make the battery feature co-exists with 0x1000 and 0x1004, allowing the device to export voltage information in addition to the battery percentage or level.

Quoted 8 lines

I tested this patch with a MX Anywhere 3, which supports the new feature. Since I don't have any device that doesn't support the state_of_charge parameter of this feature, I forced the MX Anywhere 3 to use the level information, instead of battery percentage, to test that part of the implementation.
I also tested with a MX Master 3, which supports the Battery Level Status (0x1000) feature, and a G703 Hero, which supports the Battery Voltage (0x1001) feature, to make sure nothing broke there.

Thanks a lot for the patch, Filipe. Minor details:

Quoted 27 lines
---
 drivers/hid/hid-logitech-hidpp.c | 244 ++++++++++++++++++++++++++++++-
 1 file changed, 237 insertions(+), 7 deletions(-)
drivers/hid/hid-logitech-hidpp.c +3 -0
index f85781464807..291c6b4d26b7 100644
@@ -92,6 +92,8 @@ MODULE_PARM_DESC(disable_tap_to_click,
 #define HIDPP_CAPABILITY_BATTERY_MILEAGE	BIT(2)
 #define HIDPP_CAPABILITY_BATTERY_LEVEL_STATUS	BIT(3)
 #define HIDPP_CAPABILITY_BATTERY_VOLTAGE	BIT(4)
+#define HIDPP_CAPABILITY_BATTERY_PERCENTAGE	BIT(5)
+#define HIDPP_CAPABILITY_UNIFIED_BATTERY	BIT(6)
 
 #define lg_map_key_clear(c)  hid_map_usage_clear(hi, usage, bit, max, EV_KEY, (c))
 
@@ -152,6 +154,7 @@ struct hidpp_battery {
 	int voltage;
 	int charge_type;
 	bool online;
+	u8 supported_levels_1004;
 };
 
 /**

drivers/hid/hid-logitech-hidpp.c static int hidpp20_batterylevel_get_battery_info(struct hidpp_device *hidpp, · line 1177

Quoted 5 lines
drivers/hid/hid-logitech-hidpp.c +1 -1
@@ -1171,7 +1174,7 @@ static int hidpp20_batterylevel_get_battery_info(struct hidpp_device *hidpp,
 	return 0;
 }
 
-static int hidpp20_query_battery_info(struct hidpp_device *hidpp)
+static int hidpp20_query_battery_info_1000(struct hidpp_device *hidpp)

That '_1000' suffix looks strange to me, as it's not completely obvious just from looking at the code what it actually means. Would it perhaps be more readable to call it something like hidpp20_query_battery_level(), and symmentrically change hidpp20_query_battery_info_1004() to e.g. hidpp20_query_battery_voltage() ?

[ ... snip ... ]

drivers/hid/hid-logitech-hidpp.c static int hidpp20_battery_voltage_event(struct hidpp_device *hidpp, · line 1436

Quoted 5 lines
drivers/hid/hid-logitech-hidpp.c +5 -0
+	/* if the device supports state of charge (battery percentage) we won't
+	   export the battery level information. there are 4 possible battery
+	   levels and they all are optional, this means that the device might
+	   not support any of them, we are just better off with the battery
+	   percentage. */

Could you please use standard kernel commenting style here?

Thanks,

Signature · 2 lines
Jiri Kosina
SUSE Labs
#5 lore raw
Quoted 70 lines · Jiri Kosina
Quoted 24 lines · On Mon, 4 Jan 2021, [email protected] wrote:

From: Filipe Laíns <[email protected]>

This new feature present in new devices replaces the old Battery Level Status (0x1000) feature. It keeps essentially the same information for levels (reporting critical, low, good and full) but makes these levels optional, the device exports a capability setting which describes which levels it supports. In addition to this, there is an optional state_of_charge paramenter that exports the battery percentage.

This patch adds support for this new feature. There were some implementation choices, as described below and in the code.

If the device supports the state_of_charge parameter, we will just export the battery percentage and not the levels, which the device might still support.

Since this feature can co-exist with the Battery Voltage (0x1001) feature and we currently only support one battery feature, I changed the battery feature discovery to try to use 0x1000 and 0x1004 first and only then 0x1001, the battery voltage feature.
In the future we could uncouple this and make the battery feature co-exists with 0x1000 and 0x1004, allowing the device to export voltage information in addition to the battery percentage or level.

Quoted 8 lines

I tested this patch with a MX Anywhere 3, which supports the new feature. Since I don't have any device that doesn't support the state_of_charge parameter of this feature, I forced the MX Anywhere 3 to use the level information, instead of battery percentage, to test that part of the implementation.
I also tested with a MX Master 3, which supports the Battery Level Status (0x1000) feature, and a G703 Hero, which supports the Battery Voltage (0x1001) feature, to make sure nothing broke there.

Thanks a lot for the patch, Filipe. Minor details:

Quoted 33 lines

---
 drivers/hid/hid-logitech-hidpp.c | 244 ++++++++++++++++++++++++++++++-

 1 file changed, 237 insertions(+), 7 deletions(-)
drivers/hid/hid-logitech- renamed from drivers/hid/hid-logitech-hidpp.c +0 -0

hidpp.c
index f85781464807..291c6b4d26b7 100644

drivers/hid/hid-logitech-hidpp.c +2 -0
@@ -92,6 +92,8 @@ MODULE_PARM_DESC(disable_tap_to_click,
  #define HIDPP_CAPABILITY_BATTERY_MILEAGE       BIT(2)
  #define HIDPP_CAPABILITY_BATTERY_LEVEL_STATUS  BIT(3)
  #define HIDPP_CAPABILITY_BATTERY_VOLTAGE       BIT(4)
+#define HIDPP_CAPABILITY_BATTERY_PERCENTAGE    BIT(5)
+#define HIDPP_CAPABILITY_UNIFIED_BATTERY       BIT(6)


 #define lg_map_key_clear(c)  hid_map_usage_clear(hi, usage, bit, max, EV_KEY, (c))

drivers/hid/hid-logitech-hidpp.c +1 -0
@@ -152,6 +154,7 @@ struct hidpp_battery {
         int voltage;
         int charge_type;
         bool online;
+       u8 supported_levels_1004;
  };


 /**

drivers/hid/hid-logitech-hidpp.c +0 -0
@@ -1171,7 +1174,7 @@ static int
 hidpp20_batterylevel_get_battery_info(struct hidpp_device *hidpp,
         return 0;
  }

Quoted 8 lines

drivers/hid/hid-logitech-hidpp.c static int hidpp20_batterylevel_get_battery_info(struct hidpp_device *hidpp, · line 1177

Quoted 2 lines
drivers/hid/hid-logitech-hidpp.c +1 -1
-static int hidpp20_query_battery_info(struct hidpp_device *hidpp)
+static int hidpp20_query_battery_info_1000(struct hidpp_device *hidpp)

That '_1000' suffix looks strange to me, as it's not completely obvious just from looking at the code what it actually means. Would it perhaps be more readable to call it something like hidpp20_query_battery_level(), and symmentrically change hidpp20_query_battery_info_1004() to e.g. hidpp20_query_battery_voltage() ?

The problem here is that hidpp20_query_battery_info_1004() does not set the battery voltage, it is also the battery level. The best alternative I can think of is replacing the 1000/1004 with slightly mangled HID++ feature names, like we do on the other feature function. The drawback here is that I think that could get confusing quickly.

hidpp20_batterylevel_query_battery_info()
hidpp20_unifiedbattery_query_battery_info()

Note that this does not provide *that* much more information than the feature number, though it is probably the best option. What do you think?

Quoted 11 lines

[ ... snip ... ]

Quoted 8 lines

+       /* if the device supports state of charge (battery percentage) we won't
+          export the battery level information. there are 4 possible battery
+          levels and they all are optional, this means that the device might
+          not support any of them, we are just better off with the battery +          percentage. */

Could you please use standard kernel commenting style here?

Oops, sorry. Will do :)

Cheers,
Filipe Laíns

maintainer
#6 lore raw
Quoted 15 lines · Filipe Laíns
Quoted 8 lines

drivers/hid/hid-logitech-hidpp.c static int hidpp20_batterylevel_get_battery_info(struct hidpp_device *hidpp, · line 1177

Quoted 2 lines
drivers/hid/hid-logitech-hidpp.c +1 -1
-static int hidpp20_query_battery_info(struct hidpp_device *hidpp)
+static int hidpp20_query_battery_info_1000(struct hidpp_device *hidpp)

That '_1000' suffix looks strange to me, as it's not completely obvious just from looking at the code what it actually means. Would it perhaps be more readable to call it something like hidpp20_query_battery_level(), and symmentrically change hidpp20_query_battery_info_1004() to e.g. hidpp20_query_battery_voltage() ?

The problem here is that hidpp20_query_battery_info_1004() does not set the battery voltage, it is also the battery level. The best alternative I can think of is replacing the 1000/1004 with slightly mangled HID++ feature names, like we do on the other feature function. The drawback here is that I think that could get confusing quickly.

Quoted 5 lines

hidpp20_batterylevel_query_battery_info()
hidpp20_unifiedbattery_query_battery_info()

Note that this does not provide *that* much more information than the feature number, though it is probably the best option. What do you think?

Alright, what a mess :) Would it perhaps help if there is at least a short comment preceding the function definition, noting what the constants actually are?

Quoted 10 lines
Quoted 10 lines

[ ... snip ... ]

Quoted 8 lines

+       /* if the device supports state of charge (battery percentage) we won't
+          export the battery level information. there are 4 possible battery
+          levels and they all are optional, this means that the device might
+          not support any of them, we are just better off with the battery +          percentage. */

Quoted 3 lines
Quoted 1 line

Could you please use standard kernel commenting style here?

Oops, sorry. Will do :)

Thanks,

Signature · 2 lines
Jiri Kosina
SUSE Labs
#7 lore raw
Quoted 13 lines · Jiri Kosina
Quoted 12 lines · Filipe Laíns

The problem here is that hidpp20_query_battery_info_1004() does not set battery voltage, it is also the battery level. The best alternative Ican think
of is replacing the 1000/1004 with slightly mangled HID++ feature names, like we
do on the other feature function. The drawback here is that I think that could
get confusing quickly.

hidpp20_batterylevel_query_battery_info()
hidpp20_unifiedbattery_query_battery_info()

Quoted 7 lines
Quoted 3 lines

Note that this does not provide *that* much more information than the feature
number, though it is probably the best option. What do you think?

Alright, what a mess :) Would it perhaps help if there is at least a short comment preceding the function definition, noting what the constants actually are?

Yeah :head_scratch:
There is a header comment at the start of each feature section, which I think does a good job pointing this out. IMO the problem with the naming is more for people who see its usage in other parts of the code, but I guess that is C for you right? Names don't scale well with code quantity :P

/* -------------------------------------------------------------------------- */
/* 0x1000: Battery level status */
/* -------------------------------------------------------------------------- */

/* -------------------------------------------------------------------------- */
/* 0x1004: Unified battery */
/* -------------------------------------------------------------------------- */

Quoted 6 lines
Quoted 3 lines
Quoted 1 line

Could you please use standard kernel commenting style here?

Oops, sorry. Will do :)

Thanks,

Cheers,
Filipe Laíns

maintainer
#8 lore raw
Quoted 5 lines · Filipe Laíns

Yeah :head_scratch:
There is a header comment at the start of each feature section, which I think does a good job pointing this out. IMO the problem with the naming is more for people who see its usage in other parts of the code, but I guess that is C for you right? Names don't scale well with code quantity :P

Alright ... thug life :) Let's just keep it the way it is.

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 1 line

Could you please use standard kernel commenting style here?

Oops, sorry. Will do :)

I have adjusted the wrong comment and applied. Thanks,

Signature · 2 lines
Jiri Kosina
SUSE Labs