| Message ID | 20241111-p3t1085-v3-3-bff511550aad@nxp.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | hwmon: tmp108: Add support for P3T1085 | expand |
On 11/11/24 09:32, Frank Li wrote: > Add help function tmp108_common_probe() to pave road to support i3c for > P3T1085(NXP) chip. > > Use dev_err_probe() to simple code. > > Signed-off-by: Frank Li <Frank.Li@nxp.com> I applied the first two patches of the series, but I am holding back this and the subsequent patches since I don't know how the kernel handles having two drivers in a single module, and one of them fails to instantiate. I did not find a matching example elsewhere in the kernel. We may have to use the method used by the st_lsm6dsx driver - essentially splitting the driver into three parts (i2c, i3c, and common). That would be overkill; the alternative might be something like the approach used by module_i3c_i2c_driver(), but with a set of #ifdefs around (some of) the i3c code. Guenter
On Mon, Nov 11, 2024 at 11:44:08AM -0800, Guenter Roeck wrote: > On 11/11/24 09:32, Frank Li wrote: > > Add help function tmp108_common_probe() to pave road to support i3c for > > P3T1085(NXP) chip. > > > > Use dev_err_probe() to simple code. > > > > Signed-off-by: Frank Li <Frank.Li@nxp.com> > > I applied the first two patches of the series, but I am holding back this > and the subsequent patches since I don't know how the kernel handles having > two drivers in a single module, and one of them fails to instantiate. I did > not find a matching example elsewhere in the kernel. We may have to use > the method used by the st_lsm6dsx driver - essentially splitting the driver > into three parts (i2c, i3c, and common). That would be overkill; the alternative > might be something like the approach used by module_i3c_i2c_driver(), but with > a set of #ifdefs around (some of) the i3c code. Thanks, let me think more about this. > > Guenter >
diff --git a/drivers/hwmon/tmp108.c b/drivers/hwmon/tmp108.c index b561b452d8d39..bfbea6349a95f 100644 --- a/drivers/hwmon/tmp108.c +++ b/drivers/hwmon/tmp108.c @@ -323,33 +323,19 @@ static const struct regmap_config tmp108_regmap_config = { .use_single_write = true, }; -static int tmp108_probe(struct i2c_client *client) +static int tmp108_common_probe(struct device *dev, struct regmap *regmap, char *name) { - struct device *dev = &client->dev; struct device *hwmon_dev; struct tmp108 *tmp108; - int err; u32 config; - - if (!i2c_check_functionality(client->adapter, - I2C_FUNC_SMBUS_WORD_DATA)) { - dev_err(dev, - "adapter doesn't support SMBus word transactions\n"); - return -ENODEV; - } + int err; tmp108 = devm_kzalloc(dev, sizeof(*tmp108), GFP_KERNEL); if (!tmp108) return -ENOMEM; dev_set_drvdata(dev, tmp108); - - tmp108->regmap = devm_regmap_init_i2c(client, &tmp108_regmap_config); - if (IS_ERR(tmp108->regmap)) { - err = PTR_ERR(tmp108->regmap); - dev_err(dev, "regmap init failed: %d", err); - return err; - } + tmp108->regmap = regmap; err = regmap_read(tmp108->regmap, TMP108_REG_CONF, &config); if (err < 0) { @@ -383,13 +369,30 @@ static int tmp108_probe(struct i2c_client *client) return err; } - hwmon_dev = devm_hwmon_device_register_with_info(dev, client->name, + hwmon_dev = devm_hwmon_device_register_with_info(dev, name, tmp108, &tmp108_chip_info, NULL); return PTR_ERR_OR_ZERO(hwmon_dev); } +static int tmp108_probe(struct i2c_client *client) +{ + struct device *dev = &client->dev; + struct regmap *regmap; + + if (!i2c_check_functionality(client->adapter, + I2C_FUNC_SMBUS_WORD_DATA)) + return dev_err_probe(dev, -ENODEV, + "adapter doesn't support SMBus word transactions\n"); + + regmap = devm_regmap_init_i2c(client, &tmp108_regmap_config); + if (IS_ERR(regmap)) + return dev_err_probe(dev, PTR_ERR(regmap), "regmap init failed"); + + return tmp108_common_probe(dev, regmap, client->name); +} + static int tmp108_suspend(struct device *dev) { struct tmp108 *tmp108 = dev_get_drvdata(dev);
Add help function tmp108_common_probe() to pave road to support i3c for P3T1085(NXP) chip. Use dev_err_probe() to simple code. Signed-off-by: Frank Li <Frank.Li@nxp.com> --- dev_err_probe() have not involve addition diff change. The difference always list these code block change regardless use dev_err_probe(). --- change from v2 to v3 - update subject by add prepare I3C support" --- drivers/hwmon/tmp108.c | 39 +++++++++++++++++++++------------------ 1 file changed, 21 insertions(+), 18 deletions(-)