On Thu, Jul 07, 2016 at 09:39:00AM -0700, jeffrey.lin wrote:
> ts->reset_gpio=0 will let raydium IC enter reset mode, and ts->reset_gpio=1
> will be normal touch mode.
>
> Signed-off-by: jeffrey.lin <jeffrey....@rad-ic.com>
As I mentioned in the original review, this is wrong. gpiod APIs take
into account the declared polarity of a given line, so 1 always means
"active" and 0 means "inactive", even for "active low" pins.
> ---
> drivers/input/touchscreen/raydium_i2c_ts.c | 9 ++++-----
> 1 file changed, 4 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/input/touchscreen/raydium_i2c_ts.c
> b/drivers/input/touchscreen/raydium_i2c_ts.c
> index f3076d9..5217339 100644
> --- a/drivers/input/touchscreen/raydium_i2c_ts.c
> +++ b/drivers/input/touchscreen/raydium_i2c_ts.c
> @@ -78,7 +78,7 @@
> #define RM_MAX_FW_RETRIES 30
> #define RM_MAX_FW_SIZE 0xD000
>
> -#define RM_POWERON_DELAY_USEC 500
> +#define RM_POWERON_DELAY_MSEC 20
Why do we need to increase this delay?
> #define RM_RESET_DELAY_MSEC 50
>
> enum raydium_bl_cmd {
> @@ -933,7 +933,7 @@ static int raydium_i2c_power_on(struct raydium_data *ts)
> if (!ts->reset_gpio)
> return 0;
>
> - gpiod_set_value_cansleep(ts->reset_gpio, 1);
> + gpiod_set_value_cansleep(ts->reset_gpio, 0);
>
> error = regulator_enable(ts->avdd);
> if (error) {
> @@ -950,10 +950,10 @@ static int raydium_i2c_power_on(struct raydium_data *ts)
> goto release_reset_gpio;
> }
>
> - udelay(RM_POWERON_DELAY_USEC);
> + msleep(RM_POWERON_DELAY_MSEC);
>
> release_reset_gpio:
> - gpiod_set_value_cansleep(ts->reset_gpio, 0);
> + gpiod_set_value_cansleep(ts->reset_gpio, 1);
>
> if (error)
> return error;
> @@ -968,7 +968,6 @@ static void raydium_i2c_power_off(void *_data)
> struct raydium_data *ts = _data;
>
> if (ts->reset_gpio) {
> - gpiod_set_value_cansleep(ts->reset_gpio, 1);
> regulator_disable(ts->vccio);
> regulator_disable(ts->avdd);
> }
> --
> 2.1.2
>
Thanks.
--
Dmitry
