Re: [PATCH v4 3/4] platform: arm64: Add Acer Aspire 1 embedded controller driver

[Nikita Travkin]

Bryan O'Donoghue писал(а) 12.03.2024 16:58:
> On 12/03/2024 08:42, Nikita Travkin wrote:
>> Acer Aspire 1 is a Snapdragon 7c based laptop. It uses an embedded
>> controller to perform a set of various functions, such as:
>>
>> - Battery and charger monitoring;
>> - Keyboard layout control (i.e. fn_lock settings);
>> - USB Type-C DP alt mode HPD notifications;
>> - Laptop lid status.
>>
>> Unfortunately, while all this functionality is implemented in ACPI, it's
>> currently not possible to use ACPI to boot Linux on such Qualcomm
>> devices. To allow Linux to still support the features provided by EC,
>> this driver reimplments the relevant ACPI parts. This allows us to boot
>> the laptop with Device Tree and retain all the features.
>>
>> Signed-off-by: Nikita Travkin <nikita@xxxxxxx>
>> ---
>>   drivers/platform/arm64/Kconfig           |  16 +
>>   drivers/platform/arm64/Makefile          |   2 +
>>   drivers/platform/arm64/acer-aspire1-ec.c | 555 +++++++++++++++++++++++++++++++
> 
> You should be listing yourself as a maintainer for a driver you contribute.

I always believed that being in the AUTHOR() at the bottom of the driver
would guarantee me being in CC for patches, which so far worked great,
thus I was always hesitent adding extra entries in MAINTAINERS.

> 
> Good work BTW.
> 
>>   3 files changed, 573 insertions(+)
>>
>> diff --git a/drivers/platform/arm64/Kconfig b/drivers/platform/arm64/Kconfig
>> index 644b83ede093..07d47879a9e3 100644
>> --- a/drivers/platform/arm64/Kconfig
>> +++ b/drivers/platform/arm64/Kconfig
>> @@ -16,4 +16,20 @@ menuconfig ARM64_PLATFORM_DEVICES
>>     if ARM64_PLATFORM_DEVICES
>>   +config EC_ACER_ASPIRE1
>> +	tristate "Acer Aspire 1 Emedded Controller driver"
>> +	depends on I2C
>> +	depends on DRM
>> +	depends on POWER_SUPPLY
>> +	depends on INPUT
>> +	help
>> +	  Say Y here to enable the EC driver for the (Snapdragon-based)
>> +	  Acer Aspire 1 laptop. The EC handles battery and charging
>> +	  monitoring as well as some misc functions like the lid sensor
>> +	  and USB Type-C DP HPD events.
>> +
>> +	  This driver provides battery and AC status support for the mentioned
>> +	  laptop where this information is not properly exposed via the
>> +	  standard ACPI devices.
>> +
>>   endif # ARM64_PLATFORM_DEVICES
>> diff --git a/drivers/platform/arm64/Makefile b/drivers/platform/arm64/Makefile
>> index f91cdc7155e2..4fcc9855579b 100644
>> --- a/drivers/platform/arm64/Makefile
>> +++ b/drivers/platform/arm64/Makefile
>> @@ -4,3 +4,5 @@
>>   #
>>   # This dir should only include drivers for EC-like devices.
>>   #
>> +
>> +obj-$(CONFIG_EC_ACER_ASPIRE1)	+= acer-aspire1-ec.o
>> diff --git a/drivers/platform/arm64/acer-aspire1-ec.c b/drivers/platform/arm64/acer-aspire1-ec.c
>> new file mode 100644
>> index 000000000000..3941e24c5c7c
>> --- /dev/null
>> +++ b/drivers/platform/arm64/acer-aspire1-ec.c
>> @@ -0,0 +1,555 @@
>> +// SPDX-License-Identifier: GPL-2.0-only
>> +/* Copyright (c) 2023, Nikita Travkin <nikita@xxxxxxx> */
> 
> 2024
> 

That driver was looong in the oven :D

>> +
>> +#include <linux/module.h>
>> +#include <linux/platform_device.h>
>> +#include <linux/power_supply.h>
>> +#include <linux/i2c.h>
>> +#include <linux/input.h>
>> +#include <linux/delay.h>
> 
> This list should be alphabetised.
> 
>> +
>> +#include <linux/usb/typec_mux.h>
>> +#include <drm/drm_bridge.h>
> 
> Zap the newline and sort all of these headers alphabetically please.
> 

Ack, will cleanup.

>> +
>> +#define ASPIRE_EC_EVENT			0x05
>> +
>> +#define ASPIRE_EC_EVENT_WATCHDOG	0x20
>> +#define ASPIRE_EC_EVENT_KBD_BKL_ON	0x57
>> +#define ASPIRE_EC_EVENT_KBD_BKL_OFF	0x58
>> +#define ASPIRE_EC_EVENT_LID_CLOSE	0x9b
>> +#define ASPIRE_EC_EVENT_LID_OPEN	0x9c
>> +#define ASPIRE_EC_EVENT_BKL_UNBLANKED	0x9d
>> +#define ASPIRE_EC_EVENT_BKL_BLANKED	0x9e
>> +#define ASPIRE_EC_EVENT_FG_INF_CHG	0x85
>> +#define ASPIRE_EC_EVENT_FG_STA_CHG	0xc6
>> +#define ASPIRE_EC_EVENT_HPD_DIS		0xa3
>> +#define ASPIRE_EC_EVENT_HPD_CON		0xa4
>> +
>> +#define ASPIRE_EC_FG_DYNAMIC		0x07
>> +#define ASPIRE_EC_FG_STATIC		0x08
>> +
>> +#define ASPIRE_EC_FG_FLAG_PRESENT	BIT(0)
>> +#define ASPIRE_EC_FG_FLAG_FULL		BIT(1)
>> +#define ASPIRE_EC_FG_FLAG_DISCHARGING	BIT(2)
>> +#define ASPIRE_EC_FG_FLAG_CHARGING	BIT(3)
>> +
>> +#define ASPIRE_EC_RAM_READ		0x20
>> +#define ASPIRE_EC_RAM_WRITE		0x21
>> +
>> +#define ASPIRE_EC_RAM_WATCHDOG		0x19
>> +#define ASPIRE_EC_RAM_KBD_MODE		0x43
>> +
>> +#define ASPIRE_EC_RAM_KBD_FN_EN		BIT(0)
>> +#define ASPIRE_EC_RAM_KBD_MEDIA_ON_TOP	BIT(5)
>> +#define ASPIRE_EC_RAM_KBD_ALWAYS_SET	BIT(6)
>> +#define ASPIRE_EC_RAM_KBD_NUM_LAYER_EN	BIT(7)
>> +
>> +#define ASPIRE_EC_RAM_KBD_MODE_2	0x60
>> +
>> +#define ASPIRE_EC_RAM_KBD_MEDIA_NOTIFY	BIT(3)
>> +
>> +#define ASPIRE_EC_RAM_HPD_STATUS	0xf4
>> +#define ASPIRE_EC_HPD_CONNECTED		0x03
>> +
>> +#define ASPIRE_EC_RAM_LID_STATUS	0x4c
>> +#define ASPIRE_EC_LID_OPEN		BIT(6)
>> +
>> +#define ASPIRE_EC_RAM_ADP		0x40
>> +#define ASPIRE_EC_AC_STATUS		BIT(0)
>> +
>> +struct aspire_ec {
>> +	struct i2c_client *client;
>> +	struct power_supply *bat_psy;
>> +	struct power_supply *adp_psy;
>> +	struct input_dev *idev;
>> +
>> +	bool bridge_configured;
>> +	struct drm_bridge bridge;
>> +	struct work_struct work;
>> +};
>> +
>> +static int aspire_ec_ram_read(struct i2c_client *client, u8 off, u8 *data, u8 data_len)
>> +{
>> +	i2c_smbus_write_byte_data(client, ASPIRE_EC_RAM_READ, off);
>> +	i2c_smbus_read_i2c_block_data(client, ASPIRE_EC_RAM_READ, data_len, data);
>> +	return 0;
>> +}
>> +
>> +static int aspire_ec_ram_write(struct i2c_client *client, u8 off, u8 data)
>> +{
>> +	u8 tmp[2] = {off, data};
>> +
>> +	i2c_smbus_write_i2c_block_data(client, ASPIRE_EC_RAM_WRITE, sizeof(tmp), tmp);
>> +	return 0;
>> +}
>> +
>> +static irqreturn_t aspire_ec_irq_handler(int irq, void *data)
>> +{
>> +	struct aspire_ec *ec = data;
>> +	int id;
>> +	u8 tmp;
>> +
>> +	/*
>> +	 * The original ACPI firmware actually has a small sleep in the handler.
>> +	 *
>> +	 * It seems like in most cases it's not needed but when the device
>> +	 * just exits suspend, our i2c driver has a brief time where data
>> +	 * transfer is not possible yet. So this delay allows us to suppress
>> +	 * quite a bunch of spurious error messages in dmesg. Thus it's kept.
>> +	 */
>> +	usleep_range(15000, 30000);
>> +
>> +	id = i2c_smbus_read_byte_data(ec->client, ASPIRE_EC_EVENT);
>> +	if (id < 0) {
>> +		dev_err(&ec->client->dev, "Failed to read event id: %pe\n", ERR_PTR(id));
>> +		return IRQ_HANDLED;
>> +	}
>> +
>> +	switch (id) {
>> +	case 0x0: /* No event */
>> +		break;
>> +
>> +	case ASPIRE_EC_EVENT_WATCHDOG:
>> +		/*
>> +		 * Here acpi responds to the event and clears some bit.
>> +		 * Notify (\_SB.I2C3.BAT1, 0x81) // Information Change
>> +		 * Notify (\_SB.I2C3.ADP1, 0x80) // Status Change
>> +		 */
>> +		aspire_ec_ram_read(ec->client, ASPIRE_EC_RAM_WATCHDOG, &tmp, sizeof(tmp));
>> +		aspire_ec_ram_write(ec->client, ASPIRE_EC_RAM_WATCHDOG, tmp & 0xbf);
>> +		break;
>> +
>> +	case ASPIRE_EC_EVENT_LID_CLOSE:
>> +		/* Notify (\_SB.LID0, 0x80) // Status Change */
>> +		input_report_switch(ec->idev, SW_LID, 1);
>> +		input_sync(ec->idev);
>> +		break;
>> +
>> +	case ASPIRE_EC_EVENT_LID_OPEN:
>> +		/* Notify (\_SB.LID0, 0x80) // Status Change */
>> +		input_report_switch(ec->idev, SW_LID, 0);
>> +		input_sync(ec->idev);
>> +		break;
>> +
>> +	case ASPIRE_EC_EVENT_FG_INF_CHG:
>> +		/* Notify (\_SB.I2C3.BAT1, 0x81) // Information Change */
> 
> fallthrough;
> 

Hm I believe this would not warn since it's just two values for the same
code, just with an extra comment inbetween?

>> +	case ASPIRE_EC_EVENT_FG_STA_CHG:
>> +		/* Notify (\_SB.I2C3.BAT1, 0x80) // Status Change */
>> +		power_supply_changed(ec->bat_psy);
>> +		power_supply_changed(ec->adp_psy);
>> +		break;
>> +
>> +	case ASPIRE_EC_EVENT_HPD_DIS:
>> +		if (ec->bridge_configured)
>> +			drm_bridge_hpd_notify(&ec->bridge, connector_status_disconnected);
>> +		break;
>> +
>> +	case ASPIRE_EC_EVENT_HPD_CON:
>> +		if (ec->bridge_configured)
>> +			drm_bridge_hpd_notify(&ec->bridge, connector_status_connected);
>> +		break;
>> +
>> +	case ASPIRE_EC_EVENT_BKL_BLANKED:
>> +	case ASPIRE_EC_EVENT_BKL_UNBLANKED:
>> +		/* Display backlight blanked on FN+F6. No action needed. */
>> +		break;
>> +
>> +	case ASPIRE_EC_EVENT_KBD_BKL_ON:
>> +	case ASPIRE_EC_EVENT_KBD_BKL_OFF:
>> +		/*
>> +		 * There is a keyboard backlight connector on Aspire 1 that is
>> +		 * controlled by FN+F8. There is no kb backlight on the device though.
>> +		 * Seems like this is used on other devices like Acer Spin 7.
>> +		 * No action needed.
>> +		 */
>> +		break;
>> +
>> +	default:
>> +		dev_warn(&ec->client->dev, "Unknown event id=0x%x\n", id);
>> +	}
>> +
>> +	return IRQ_HANDLED;
>> +}
>> +
>> +/*
>> + * Power supply.
>> + */
>> +
>> +struct aspire_ec_bat_psy_static_data {
>> +	u8 unk1;
>> +	u8 flags;
>> +	__le16 unk2;
>> +	__le16 voltage_design;
>> +	__le16 capacity_full;
>> +	__le16 unk3;
>> +	__le16 serial;
>> +	u8 model_id;
>> +	u8 vendor_id;
>> +} __packed;
>> +
>> +static const char * const aspire_ec_bat_psy_battery_model[] = {
>> +	"AP18C4K",
>> +	"AP18C8K",
>> +	"AP19B8K",
>> +	"AP16M4J",
>> +	"AP16M5J",
>> +};
>> +
>> +static const char * const aspire_ec_bat_psy_battery_vendor[] = {
>> +	"SANYO",
>> +	"SONY",
>> +	"PANASONIC",
>> +	"SAMSUNG",
>> +	"SIMPLO",
>> +	"MOTOROLA",
>> +	"CELXPERT",
>> +	"LGC",
>> +	"GETAC",
>> +	"MURATA",
>> +};
>> +
>> +struct aspire_ec_bat_psy_dynamic_data {
>> +	u8 unk1;
>> +	u8 flags;
>> +	u8 unk2;
>> +	__le16 capacity_now;
>> +	__le16 voltage_now;
>> +	__le16 current_now;
>> +	__le16 unk3;
>> +	__le16 unk4;
>> +} __packed;
>> +
>> +static int aspire_ec_bat_psy_get_property(struct power_supply *psy,
>> +				      enum power_supply_property psp,
>> +				      union power_supply_propval *val)
>> +{
>> +	struct aspire_ec *ec = power_supply_get_drvdata(psy);
>> +	struct aspire_ec_bat_psy_static_data sdat;
>> +	struct aspire_ec_bat_psy_dynamic_data ddat;
>> +
>> +	i2c_smbus_read_i2c_block_data(ec->client, ASPIRE_EC_FG_STATIC, sizeof(sdat), (u8 *)&sdat);
>> +	i2c_smbus_read_i2c_block_data(ec->client, ASPIRE_EC_FG_DYNAMIC, sizeof(ddat), (u8 *)&ddat);
>> +
>> +	switch (psp) {
>> +	case POWER_SUPPLY_PROP_STATUS:
>> +		val->intval = POWER_SUPPLY_STATUS_UNKNOWN;
>> +		if (ddat.flags & ASPIRE_EC_FG_FLAG_CHARGING)
>> +			val->intval = POWER_SUPPLY_STATUS_CHARGING;
>> +		else if (ddat.flags & ASPIRE_EC_FG_FLAG_DISCHARGING)
>> +			val->intval = POWER_SUPPLY_STATUS_DISCHARGING;
>> +		else if (ddat.flags & ASPIRE_EC_FG_FLAG_FULL)
>> +			val->intval = POWER_SUPPLY_STATUS_FULL;
>> +		break;
>> +
>> +	case POWER_SUPPLY_PROP_VOLTAGE_NOW:
>> +		val->intval = le16_to_cpu(ddat.voltage_now) * 1000;
>> +		break;
>> +
>> +	case POWER_SUPPLY_PROP_VOLTAGE_MAX_DESIGN:
>> +		val->intval = le16_to_cpu(sdat.voltage_design) * 1000;
>> +		break;
>> +
>> +	case POWER_SUPPLY_PROP_CHARGE_NOW:
>> +		val->intval = le16_to_cpu(ddat.capacity_now) * 1000;
>> +		break;
>> +
>> +	case POWER_SUPPLY_PROP_CHARGE_FULL:
>> +		val->intval = le16_to_cpu(sdat.capacity_full) * 1000;
>> +		break;
> 
> You could stick this "* 1000" stuff in a macro
> 

acpi/battery.c also explicitly sets the multiplier so I think it's the
"common" way to do this.

>> +
>> +	case POWER_SUPPLY_PROP_CAPACITY:
>> +		val->intval = le16_to_cpu(ddat.capacity_now) * 100;
>> +		val->intval /= le16_to_cpu(sdat.capacity_full);
>> +		break;
>> +
>> +	case POWER_SUPPLY_PROP_CURRENT_NOW:
>> +		val->intval = (s16)le16_to_cpu(ddat.current_now) * 1000;
>> +		break;
>> +
>> +	case POWER_SUPPLY_PROP_PRESENT:
>> +		val->intval = !!(ddat.flags & ASPIRE_EC_FG_FLAG_PRESENT);
>> +		break;
>> +
>> +	case POWER_SUPPLY_PROP_SCOPE:
>> +		val->intval = POWER_SUPPLY_SCOPE_SYSTEM;
>> +		break;
>> +
>> +	case POWER_SUPPLY_PROP_MODEL_NAME:
>> +		if (sdat.model_id - 1 < ARRAY_SIZE(aspire_ec_bat_psy_battery_model))
>> +			val->strval = aspire_ec_bat_psy_battery_model[sdat.model_id - 1];
>> +		else
>> +			val->strval = "Unknown";
>> +		break;
>> +
>> +	case POWER_SUPPLY_PROP_MANUFACTURER:
>> +		if (sdat.vendor_id - 3 < ARRAY_SIZE(aspire_ec_bat_psy_battery_vendor))
>> +			val->strval = aspire_ec_bat_psy_battery_vendor[sdat.vendor_id - 3];
> 
> How does this -3 offset not underflow ?
> 

vendor_id here is unsigned so the if check would actually overflow,
though explaining that I guess it's better to be explicit there and let
the compiler optimize that check away anyway... I will update the if
condition with an extra (id >= 3).

> Seems a bit dodgy to me - can you add a comment to the code to explain ? Its not immediately obvious the -3 is OK.
> 
> Also could you take an index instead of replicating the -value stepdown each time ?
> 
> int myindex = sdat.model_id - 1;
> 
> if (myindex < someconstraint)
> 	strval = somearry[myindex];
> 

I decided against adding a dedicated index variable since there is only
one actual use for each, so it's easy to see where it goes.

>> +		else
>> +			val->strval = "Unknown";
>> +		break;
>> +
>> +	default:
>> +		return -EINVAL;
>> +	}
>> +
>> +	return 0;
>> +}
>> +
>> +static enum power_supply_property aspire_ec_bat_psy_props[] = {
>> +	POWER_SUPPLY_PROP_STATUS,
>> +	POWER_SUPPLY_PROP_VOLTAGE_NOW,
>> +	POWER_SUPPLY_PROP_VOLTAGE_MAX_DESIGN,
>> +	POWER_SUPPLY_PROP_CHARGE_NOW,
>> +	POWER_SUPPLY_PROP_CHARGE_FULL,
>> +	POWER_SUPPLY_PROP_CAPACITY,
>> +	POWER_SUPPLY_PROP_CURRENT_NOW,
>> +	POWER_SUPPLY_PROP_PRESENT,
>> +	POWER_SUPPLY_PROP_SCOPE,
>> +	POWER_SUPPLY_PROP_MODEL_NAME,
>> +	POWER_SUPPLY_PROP_MANUFACTURER,
>> +};
>> +
>> +static const struct power_supply_desc aspire_ec_bat_psy_desc = {
>> +	.name		= "aspire-ec-bat",
>> +	.type		= POWER_SUPPLY_TYPE_BATTERY,
>> +	.get_property	= aspire_ec_bat_psy_get_property,
>> +	.properties	= aspire_ec_bat_psy_props,
>> +	.num_properties	= ARRAY_SIZE(aspire_ec_bat_psy_props),
>> +};
>> +
>> +static int aspire_ec_adp_psy_get_property(struct power_supply *psy,
>> +				      enum power_supply_property psp,
>> +				      union power_supply_propval *val)
>> +{
>> +	struct aspire_ec *ec = power_supply_get_drvdata(psy);
>> +	u8 tmp;
>> +
>> +	switch (psp) {
>> +	case POWER_SUPPLY_PROP_ONLINE:
>> +		aspire_ec_ram_read(ec->client, ASPIRE_EC_RAM_ADP, &tmp, sizeof(tmp));
>> +		val->intval = !!(tmp & ASPIRE_EC_AC_STATUS);
>> +		break;
>> +
>> +	default:
>> +		return -EINVAL;
>> +	}
>> +
>> +	return 0;
>> +}
>> +
>> +static enum power_supply_property aspire_ec_adp_psy_props[] = {
>> +	POWER_SUPPLY_PROP_ONLINE,
>> +};
>> +
>> +static const struct power_supply_desc aspire_ec_adp_psy_desc = {
>> +	.name		= "aspire-ec-adp",
>> +	.type		= POWER_SUPPLY_TYPE_MAINS,
>> +	.get_property	= aspire_ec_adp_psy_get_property,
>> +	.properties	= aspire_ec_adp_psy_props,
>> +	.num_properties	= ARRAY_SIZE(aspire_ec_adp_psy_props),
>> +};
>> +
>> +/*
>> + * USB-C DP Alt mode HPD.
>> + */
>> +
>> +static int aspire_ec_bridge_attach(struct drm_bridge *bridge, enum drm_bridge_attach_flags flags)
>> +{
>> +	return flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR ? 0 : -EINVAL;
>> +}
>> +
>> +static void aspire_ec_bridge_update_hpd_work(struct work_struct *work)
>> +{
>> +	struct aspire_ec *ec = container_of(work, struct aspire_ec, work);
>> +	u8 tmp;
>> +
>> +	aspire_ec_ram_read(ec->client, ASPIRE_EC_RAM_HPD_STATUS, &tmp, sizeof(tmp));
>> +	if (tmp == ASPIRE_EC_HPD_CONNECTED)
>> +		drm_bridge_hpd_notify(&ec->bridge, connector_status_connected);
>> +	else
>> +		drm_bridge_hpd_notify(&ec->bridge, connector_status_disconnected);
>> +}
>> +
>> +static void aspire_ec_bridge_hpd_enable(struct drm_bridge *bridge)
>> +{
>> +	struct aspire_ec *ec = container_of(bridge, struct aspire_ec, bridge);
>> +
>> +	schedule_work(&ec->work);
>> +}
>> +
>> +static const struct drm_bridge_funcs aspire_ec_bridge_funcs = {
>> +	.hpd_enable = aspire_ec_bridge_hpd_enable,
>> +	.attach = aspire_ec_bridge_attach,
>> +};
>> +
>> +/*
>> + * Sysfs attributes.
>> + */
>> +
>> +static ssize_t fn_lock_show(struct device *dev, struct device_attribute *attr, char *buf)
>> +{
>> +	struct aspire_ec *ec = i2c_get_clientdata(to_i2c_client(dev));
>> +	u8 tmp;
>> +
>> +	aspire_ec_ram_read(ec->client, ASPIRE_EC_RAM_KBD_MODE, &tmp, sizeof(tmp));
>> +
>> +	return sysfs_emit(buf, "%d\n", !(tmp & ASPIRE_EC_RAM_KBD_MEDIA_ON_TOP));
>> +}
>> +
>> +static ssize_t fn_lock_store(struct device *dev, struct device_attribute *attr,
>> +			     const char *buf, size_t count)
>> +{
>> +	struct aspire_ec *ec = i2c_get_clientdata(to_i2c_client(dev));
>> +	u8 tmp;
>> +
>> +	bool state;
>> +	int ret;
>> +
>> +	ret = kstrtobool(buf, &state);
>> +	if (ret)
>> +		return ret;
>> +
>> +	aspire_ec_ram_read(ec->client, ASPIRE_EC_RAM_KBD_MODE, &tmp, sizeof(tmp));
>> +
>> +	if (state)
>> +		tmp &= ~ASPIRE_EC_RAM_KBD_MEDIA_ON_TOP;
>> +	else
>> +		tmp |= ASPIRE_EC_RAM_KBD_MEDIA_ON_TOP;
>> +
>> +	aspire_ec_ram_write(ec->client, ASPIRE_EC_RAM_KBD_MODE, tmp);
>> +
>> +	return count;
>> +}
>> +
>> +static DEVICE_ATTR_RW(fn_lock);
>> +
>> +static struct attribute *aspire_ec_attributes[] = {
>> +	&dev_attr_fn_lock.attr,
>> +	NULL
>> +};
>> +
>> +static const struct attribute_group aspire_ec_attribute_group = {
>> +	.attrs = aspire_ec_attributes
>> +};
>> +
>> +static int aspire_ec_probe(struct i2c_client *client)
>> +{
>> +	struct power_supply_config psy_cfg = {0};
>> +	struct device *dev = &client->dev;
>> +	struct fwnode_handle *fwnode;
>> +	struct aspire_ec *ec;
>> +	int ret;
>> +	u8 tmp;
>> +
>> +	ec = devm_kzalloc(dev, sizeof(*ec), GFP_KERNEL);
>> +	if (!ec)
>> +		return -ENOMEM;
>> +
>> +	ec->client = client;
>> +	i2c_set_clientdata(client, ec);
>> +
>> +	/* Battery status reports */
>> +	psy_cfg.drv_data = ec;
>> +	ec->bat_psy = devm_power_supply_register(dev, &aspire_ec_bat_psy_desc, &psy_cfg);
>> +	if (IS_ERR(ec->bat_psy))
>> +		return dev_err_probe(dev, PTR_ERR(ec->bat_psy),
>> +				     "Failed to register battery power supply\n");
>> +
>> +	ec->adp_psy = devm_power_supply_register(dev, &aspire_ec_adp_psy_desc, &psy_cfg);
>> +	if (IS_ERR(ec->adp_psy))
>> +		return dev_err_probe(dev, PTR_ERR(ec->adp_psy),
>> +				     "Failed to register ac power supply\n");
>> +
>> +	/* Lid switch */
>> +	ec->idev = devm_input_allocate_device(dev);
>> +	if (!ec->idev)
>> +		return -ENOMEM;
>> +
>> +	ec->idev->name = "aspire-ec";
>> +	ec->idev->phys = "aspire-ec/input0";
>> +	input_set_capability(ec->idev, EV_SW, SW_LID);
>> +
>> +	ret = input_register_device(ec->idev);
>> +	if (ret)
>> +		return dev_err_probe(dev, ret, "Input device register failed\n");
>> +
>> +	/* Enable the keyboard fn keys */
>> +	tmp = ASPIRE_EC_RAM_KBD_FN_EN | ASPIRE_EC_RAM_KBD_ALWAYS_SET;
>> +	tmp |= ASPIRE_EC_RAM_KBD_MEDIA_ON_TOP;
>> +	aspire_ec_ram_write(client, ASPIRE_EC_RAM_KBD_MODE, tmp);
>> +
>> +	aspire_ec_ram_read(client, ASPIRE_EC_RAM_KBD_MODE_2, &tmp, sizeof(tmp));
>> +	tmp |= ASPIRE_EC_RAM_KBD_MEDIA_NOTIFY;
>> +	aspire_ec_ram_write(client, ASPIRE_EC_RAM_KBD_MODE_2, tmp);
>> +
>> +	ret = devm_device_add_group(dev, &aspire_ec_attribute_group);
>> +	if (ret)
> 
> Don't you need input_unregister_device() in this case ? suggest a goto jump label to clean up down the end of your probe() function.
> 

devm_input_allocate_device guarantees that devm will un-register it,
which I myself also forget every time and have to re-check :D

 * Managed input devices do not need to be explicitly unregistered or
 * freed as it will be done automatically when owner device unbinds from
 * its driver (or binding fails).

So this driver only uses devm resources and thus doesn't need to
clean after itself.

Thanks for your review!
Nikita

>> +		return dev_err_probe(dev, ret, "Failed to create attribute group\n");
>> +
>> +	/* External Type-C display attach reports */
>> +	fwnode = device_get_named_child_node(dev, "connector");
>> +	if (fwnode) {
>> +		INIT_WORK(&ec->work, aspire_ec_bridge_update_hpd_work);
>> +		ec->bridge.funcs = &aspire_ec_bridge_funcs;
>> +		ec->bridge.of_node = to_of_node(fwnode);
>> +		ec->bridge.ops = DRM_BRIDGE_OP_HPD;
>> +		ec->bridge.type = DRM_MODE_CONNECTOR_USB;
>> +
>> +		ret = devm_drm_bridge_add(dev, &ec->bridge);
>> +		if (ret) {
>> +			fwnode_handle_put(fwnode);
>> +			return dev_err_probe(dev, ret, "Failed to register drm bridge\n");
>> +		}
>> +
>> +		ec->bridge_configured = true;
>> +	}
>> +
>> +	ret = devm_request_threaded_irq(dev, client->irq, NULL,
>> +					aspire_ec_irq_handler, IRQF_ONESHOT,
>> +					dev_name(dev), ec);
>> +	if (ret)
>> +		return dev_err_probe(dev, ret, "Failed to request irq\n");
>> +
>> +	return 0;
> 
> jump label here for stuff that doens't just clean up
> 
>> +}
>> +
>> +static int aspire_ec_resume(struct device *dev)
>> +{
>> +	struct aspire_ec *ec = i2c_get_clientdata(to_i2c_client(dev));
>> +	u8 tmp;
>> +
>> +	aspire_ec_ram_read(ec->client, ASPIRE_EC_RAM_LID_STATUS, &tmp, sizeof(tmp));
>> +	input_report_switch(ec->idev, SW_LID, !!(tmp & ASPIRE_EC_LID_OPEN));
>> +	input_sync(ec->idev);
>> +
>> +	return 0;
>> +}
>> +
>> +static const struct i2c_device_id aspire_ec_id[] = {
>> +	{ "aspire1-ec", },
>> +	{ }
>> +};
>> +MODULE_DEVICE_TABLE(i2c, aspire_ec_id);
>> +
>> +static const struct of_device_id aspire_ec_of_match[] = {
>> +	{ .compatible = "acer,aspire1-ec", },
>> +	{ }
>> +};
>> +MODULE_DEVICE_TABLE(of, aspire_ec_of_match);
>> +
>> +static DEFINE_SIMPLE_DEV_PM_OPS(aspire_ec_pm_ops, NULL, aspire_ec_resume);
>> +
>> +static struct i2c_driver aspire_ec_driver = {
>> +	.driver = {
>> +		.name = "aspire-ec",
>> +		.of_match_table = aspire_ec_of_match,
>> +		.pm = pm_sleep_ptr(&aspire_ec_pm_ops),
>> +	},
>> +	.probe = aspire_ec_probe,
>> +	.id_table = aspire_ec_id,
>> +};
>> +module_i2c_driver(aspire_ec_driver);
> 
> Don't you need an exit function to clean up stuff that's not devm_ derived ?
> 
> input_unregister_device() ?
> 
>> +
>> +MODULE_DESCRIPTION("Acer Aspire 1 embedded controller");
>> +MODULE_AUTHOR("Nikita Travkin <nikita@xxxxxxx>");
>> +MODULE_LICENSE("GPL");
>>