Message ID | 20211222141920.1482451-2-festevam@gmail.com (mailing list archive) |
---|---|
State | Superseded |
Headers | show |
Series | [RESEND,1/2] ASoC: cs4265: Fix part number ID error message | expand |
On Wed, Dec 22, 2021 at 11:19:20AM -0300, Fabio Estevam wrote: > From: Fabio Estevam <festevam@denx.de> > > There is no need to keep 'struct gpio_desc *reset_gpio' inside > the private structure becase reset_gpio is only used inside the > probe() function. > > Move it to a local scope. > > Signed-off-by: Fabio Estevam <festevam@denx.de> > --- Patch looks fine, but I do have a slight doubt would it perhaps be better to add a remove function that puts the device back into reset, thus keeping the private member? Feels like it might be a nicer state to leave the chip in if the driver unbinds. Thanks, Charles
diff --git a/sound/soc/codecs/cs4265.c b/sound/soc/codecs/cs4265.c index b89002189a2b..88d89366f816 100644 --- a/sound/soc/codecs/cs4265.c +++ b/sound/soc/codecs/cs4265.c @@ -29,7 +29,6 @@ struct cs4265_private { struct regmap *regmap; - struct gpio_desc *reset_gpio; u8 format; u32 sysclk; }; @@ -573,6 +572,7 @@ static int cs4265_i2c_probe(struct i2c_client *i2c_client, const struct i2c_device_id *id) { struct cs4265_private *cs4265; + struct gpio_desc *reset_gpio; int ret; unsigned int devid = 0; unsigned int reg; @@ -589,14 +589,14 @@ static int cs4265_i2c_probe(struct i2c_client *i2c_client, return ret; } - cs4265->reset_gpio = devm_gpiod_get_optional(&i2c_client->dev, - "reset", GPIOD_OUT_LOW); - if (IS_ERR(cs4265->reset_gpio)) - return PTR_ERR(cs4265->reset_gpio); + reset_gpio = devm_gpiod_get_optional(&i2c_client->dev, "reset", + GPIOD_OUT_LOW); + if (IS_ERR(reset_gpio)) + return PTR_ERR(reset_gpio); - if (cs4265->reset_gpio) { + if (reset_gpio) { mdelay(1); - gpiod_set_value_cansleep(cs4265->reset_gpio, 1); + gpiod_set_value_cansleep(reset_gpio, 1); } i2c_set_clientdata(i2c_client, cs4265);