Messages in this thread | | | Date | Sat, 6 Jan 2024 14:57:57 +0100 | Subject | Re: [PATCH 1/2] power: supply: Add STC3117 fuel gauge unit driver | From | Christophe JAILLET <> |
| |
Le 06/01/2024 à 11:14, Bhavin Sharma a écrit : > Adding support for stc3117 fuel gauge driver with > minimal functionality to read voltage level currently > > Signed-off-by: Bhavin Sharma <bhavin.sharma-TXG9oGKWPboh9/7i5KS7/GGXanvQGlWp@public.gmane.org> > ---
Hi,
> drivers/power/supply/Kconfig | 7 ++ > drivers/power/supply/Makefile | 1 + > drivers/power/supply/stc3117_fuel_gauge.c | 136 ++++++++++++++++++++++ > 3 files changed, 144 insertions(+) > create mode 100644 drivers/power/supply/stc3117_fuel_gauge.c > > diff --git a/drivers/power/supply/Kconfig b/drivers/power/supply/Kconfig > index f21cb05815ec..02df55b0d5f0 100644 > --- a/drivers/power/supply/Kconfig > +++ b/drivers/power/supply/Kconfig > @@ -875,6 +875,13 @@ config FUEL_GAUGE_SC27XX > Say Y here to enable support for fuel gauge with SC27XX > PMIC chips. > > +config FUEL_GAUGE_STC3117 > + tristate "STMicroelectronics STC3117 fuel gauge driver" > + depends on I2C > + help > + Say Y here to enable support for fuel gauge with STC3117 > + PMIC chips. > + > config CHARGER_UCS1002 > tristate "Microchip UCS1002 USB Port Power Controller" > depends on I2C > diff --git a/drivers/power/supply/Makefile b/drivers/power/supply/Makefile > index 58b567278034..f8ed3e5a8882 100644 > --- a/drivers/power/supply/Makefile > +++ b/drivers/power/supply/Makefile > @@ -104,6 +104,7 @@ obj-$(CONFIG_CHARGER_CROS_USBPD) += cros_usbpd-charger.o > obj-$(CONFIG_CHARGER_CROS_PCHG) += cros_peripheral_charger.o > obj-$(CONFIG_CHARGER_SC2731) += sc2731_charger.o > obj-$(CONFIG_FUEL_GAUGE_SC27XX) += sc27xx_fuel_gauge.o > +obj-$(CONFIG_FUEL_GAUGE_STC3117) += stc3117_fuel_gauge.o > obj-$(CONFIG_CHARGER_UCS1002) += ucs1002_power.o > obj-$(CONFIG_CHARGER_BD99954) += bd99954-charger.o > obj-$(CONFIG_CHARGER_WILCO) += wilco-charger.o > diff --git a/drivers/power/supply/stc3117_fuel_gauge.c b/drivers/power/supply/stc3117_fuel_gauge.c > new file mode 100644 > index 000000000000..bb033ac54858 > --- /dev/null > +++ b/drivers/power/supply/stc3117_fuel_gauge.c > @@ -0,0 +1,136 @@ > +// SPDX-License-Identifier: GPL-2.0-only > +/* > + * stc3117_fuel_gauge.c - STMicroelectronics STC3117 Fuel Gauge Driver > + * > + * Copyright (c) 2024 Silicon Signals Pvt Ltd. > + * Author: Bhavin Sharma <bhavin.sharma-TXG9oGKWPboh9/7i5KS7/GGXanvQGlWp@public.gmane.org> > + * Hardevsinh Palaniya <hardevsinh.palaniya-TXG9oGKWPbrc8hPZMNZi9dBPR1lH4CV8@public.gmane.org> > + */ > + > +
Only 1 NL is enough.
> +#include <linux/i2c.h> > +#include <linux/i2c-dev.h> > +#include <linux/kernel.h>
Is it really needed for a small driver like this?
> +#include <linux/module.h> > +#include <linux/uaccess.h> > +#include <linux/power_supply.h>
Keep alphabetical order.
> + > +#define VOLTAGE_REG_ADDR 0x08 > +#define VOLTAGE_REG_ADDR_SIZE 1 //in bytes > +#define VOLTAGE_DATA_SIZE 2 //in bytes > +#define LSB_VALUE 2200 //in micro-volts > + > + > +static int stc3117_probe(struct i2c_client *client); > +static void stc3117_dev_remove(struct i2c_client *client);
If struct i2c_driver stc3117_i2c_driver was declared at the end of the fine, these prototypes could be removed.
> + > +static int stc3117_get_property(struct power_supply *psy, > + enum power_supply_property psp, union power_supply_propval *val); > +static int stc3117_get_batt_volt(const struct i2c_client *client);
Same here. I think that a different ordering could avoid these prototypes.
> + > +const struct i2c_client *tmp_client; > +struct power_supply *stc_sply; > + > +static const struct of_device_id stc3117_of_match[] = { > + { .compatible = "st,stc3117-fgu" }, > + {}, > +}; > + > +MODULE_DEVICE_TABLE(of, stc3117_of_match); > + > +static const struct i2c_device_id stc3117_id[] = { > + {"stc3117", 0}, > + {}, > +}; > + > +
Only 1 NL is enough.
> +MODULE_DEVICE_TABLE(i2c, stc3117_id); > + > +struct i2c_driver stc3117_i2c_driver = { > + .driver = { > + .name = "stc3117_i2c_driver", > + .owner = THIS_MODULE, > + .of_match_table = of_match_ptr(stc3117_of_match), > + }, > + .probe = stc3117_probe, > + .id_table = stc3117_id, > + .remove = stc3117_dev_remove, > +}; > + > +
Only 1 NL is enough.
> +static enum power_supply_property stc3117_battery_props[] = { > + POWER_SUPPLY_PROP_VOLTAGE_NOW, > +}; > + > +static const struct power_supply_desc stc3117_battery_desc = { > + .name = "stc3117-battery", > + .type = POWER_SUPPLY_TYPE_BATTERY, > + .get_property = stc3117_get_property, > + .properties = stc3117_battery_props, > + .num_properties = ARRAY_SIZE(stc3117_battery_props), > +}; > + > +static int stc3117_get_property(struct power_supply *psy, > + enum power_supply_property psp, union power_supply_propval *val) > +{ > + switch (psp) { > + case POWER_SUPPLY_PROP_VOLTAGE_NOW: > + val->intval = stc3117_get_batt_volt(tmp_client); > + break;
This break should be indented by 1 more tab.
> + default: > + return -EINVAL; > + } > + return 0; > +} > + > +
Only 1 NL is enough.
> +static int stc3117_get_batt_volt(const struct i2c_client *stc_client) > +{ > + int ret, volt = 0; > + char i2c_tx = VOLTAGE_REG_ADDR, i2c_rx[2] = {0};
Maybe VOLTAGE_DATA_SIZE, instead of 2?
> + > + ret = i2c_master_send(stc_client, &i2c_tx, VOLTAGE_REG_ADDR_SIZE); > + if (ret > 0) { > + > + ret = i2c_master_recv(stc_client, i2c_rx, VOLTAGE_DATA_SIZE); > + if (ret > 0) { > + > + volt = (i2c_rx[1] << 8) + i2c_rx[0]; > + volt *= LSB_VALUE; > + > + return volt; > + } > + } > + > + return ret; > +}
Writting it as: ret = i2c_master_send() if (ret < 0) return ret;
ret = i2c_master_recv(); if (ret < 0) return ret;
vols = ...
would save some indentation.
> + > +static int stc3117_probe(struct i2c_client *client) > +{ > + struct power_supply_config psy_cfg = {}; > + struct device *dev; > + > + dev = &client->dev; > + > + psy_cfg.of_node = dev->of_node; > + > + tmp_client = client; > + > + stc_sply = power_supply_register(dev, &stc3117_battery_desc, &psy_cfg);
devm_power_supply_register() ? This would remove the need of the .remove() function.
> + if (IS_ERR(stc_sply)) > + pr_err("failed to register battery\n");
dev_err()?
Maybe this error code should be propagated. What the point of this driver if this function fails?
just my 2c,
CJ
> + > + return 0; > +} > + > +static void stc3117_dev_remove(struct i2c_client *client) > +{ > + power_supply_unregister(stc_sply); > +} > + > +module_i2c_driver(stc3117_i2c_driver); > + > +MODULE_LICENSE("GPL"); > +MODULE_AUTHOR("Bhavin Sharma <bhavin.sharma-TXG9oGKWPboh9/7i5KS7/GGXanvQGlWp@public.gmane.org>"); > +MODULE_AUTHOR("Hardevsinh Palaniya <hardevsinh.palaniya-TXG9oGKWPboh9/7i5KS7/GGXanvQGlWp@public.gmane.org>"); > +MODULE_DESCRIPTION("STC3117 Fuel Gauge Driver"); > +MODULE_VERSION("v1.0"); > +
| |