If you have GPIO_LIB gpio_cansleep oopses on an invalid
gpio. So better gpio_request your pin first.
Signed-off-by: Uwe Kleine-König <[email protected]>
Cc: David Brownell <[email protected]>
Cc: Raphael Assenat <[email protected]>
Cc: Richard Purdie <[email protected]>
---
Hello,
I currently start using GPIO_LIB and don't have any chips yet. The Oops will
vanish after I will have registered the chips for my SoC's gpios, but still
this way the code is more robust.
Best regards
Uwe
drivers/leds/leds-gpio.c | 8 ++++----
1 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/leds/leds-gpio.c b/drivers/leds/leds-gpio.c
index 6c0a9c4..76ddcf3 100644
--- a/drivers/leds/leds-gpio.c
+++ b/drivers/leds/leds-gpio.c
@@ -79,6 +79,10 @@ static int gpio_led_probe(struct platform_device *pdev)
cur_led = &pdata->leds[i];
led_dat = &leds_data[i];
+ ret = gpio_request(led_dat->gpio, cur_led->name);
+ if (ret < 0)
+ goto err;
+
led_dat->cdev.name = cur_led->name;
led_dat->cdev.default_trigger = cur_led->default_trigger;
led_dat->gpio = cur_led->gpio;
@@ -87,10 +91,6 @@ static int gpio_led_probe(struct platform_device *pdev)
led_dat->cdev.brightness_set = gpio_led_set;
led_dat->cdev.brightness = LED_OFF;
- ret = gpio_request(led_dat->gpio, led_dat->cdev.name);
- if (ret < 0)
- goto err;
-
gpio_direction_output(led_dat->gpio, led_dat->active_low);
INIT_WORK(&led_dat->work, gpio_led_work);
--
1.5.4.1
Hello,
Uwe Kleine-K?nig wrote:
> diff --git a/drivers/leds/leds-gpio.c b/drivers/leds/leds-gpio.c
> index 6c0a9c4..76ddcf3 100644
> --- a/drivers/leds/leds-gpio.c
> +++ b/drivers/leds/leds-gpio.c
> @@ -79,6 +79,10 @@ static int gpio_led_probe(struct platform_device *pdev)
> cur_led = &pdata->leds[i];
> led_dat = &leds_data[i];
>
> + ret = gpio_request(led_dat->gpio, cur_led->name);
The line above must read:
ret = gpio_request(led_dat->gpio, cur_led->name);
I managed to notice that the 2nd argument needs a change but I
overlooked the first somehow.
Best regards
Uwe
--
Uwe Kleine-K?nig, Software Engineer
Digi International GmbH Branch Breisach, K?ferstrasse 8, 79206 Breisach, Germany
Tax: 315/5781/0242 / VAT: DE153662976 / Reg. Amtsgericht Dortmund HRB 13962
On Thursday 14 February 2008, Uwe Kleine-K?nig wrote:
> If you have GPIO_LIB gpio_cansleep oopses on an invalid
> gpio. So better gpio_request your pin first.
>
> Signed-off-by: Uwe Kleine-K?nig <[email protected]>
> Cc: David Brownell <[email protected]>
Acked-by: David Brownell <[email protected]>
... my bad, sorry. The "first do gpio_request(), THEN
you can test gpio_cansleep()" issue got resolved in that
direction a bit late; this is the only in-tree driver
that I know would be affected. My test platforms haven't
yet been updated to 2.6.25-rc1 ... :(
> Cc: Raphael Assenat <[email protected]>
> Cc: Richard Purdie <[email protected]>
> ---
> Hello,
>
> I currently start using GPIO_LIB and don't have any chips yet. The Oops will
> vanish after I will have registered the chips for my SoC's gpios, but still
> this way the code is more robust.
>
> Best regards
> Uwe
>
> drivers/leds/leds-gpio.c | 8 ++++----
> 1 files changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/leds/leds-gpio.c b/drivers/leds/leds-gpio.c
> index 6c0a9c4..76ddcf3 100644
> --- a/drivers/leds/leds-gpio.c
> +++ b/drivers/leds/leds-gpio.c
> @@ -79,6 +79,10 @@ static int gpio_led_probe(struct platform_device *pdev)
> cur_led = &pdata->leds[i];
> led_dat = &leds_data[i];
>
> + ret = gpio_request(led_dat->gpio, cur_led->name);
> + if (ret < 0)
> + goto err;
> +
> led_dat->cdev.name = cur_led->name;
> led_dat->cdev.default_trigger = cur_led->default_trigger;
> led_dat->gpio = cur_led->gpio;
> @@ -87,10 +91,6 @@ static int gpio_led_probe(struct platform_device *pdev)
> led_dat->cdev.brightness_set = gpio_led_set;
> led_dat->cdev.brightness = LED_OFF;
>
> - ret = gpio_request(led_dat->gpio, led_dat->cdev.name);
> - if (ret < 0)
> - goto err;
> -
> gpio_direction_output(led_dat->gpio, led_dat->active_low);
>
> INIT_WORK(&led_dat->work, gpio_led_work);
> --
> 1.5.4.1
>
If you have GPIO_LIB gpio_cansleep oopses on an invalid
gpio. So better gpio_request your pin first.
Signed-off-by: Uwe Kleine-König <[email protected]>
Acked-by: David Brownell <[email protected]>
Cc: Raphael Assenat <[email protected]>
Cc: Richard Purdie <[email protected]>
---
Hello,
I already sent that patch on Thu, 14 Feb 2008. Since then I changed
gpio_request(led_dat->gpio, ...
to
gpio_request(cur_led->gpio, ...
because at that time led_dat->gpio isn't initialized yet.
And I got the Acked-by: David.
Best regards
Uwe
---
drivers/leds/leds-gpio.c | 8 ++++----
1 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/leds/leds-gpio.c b/drivers/leds/leds-gpio.c
index 6c0a9c4..d13b622 100644
--- a/drivers/leds/leds-gpio.c
+++ b/drivers/leds/leds-gpio.c
@@ -79,6 +79,10 @@ static int gpio_led_probe(struct platform_device *pdev)
cur_led = &pdata->leds[i];
led_dat = &leds_data[i];
+ ret = gpio_request(cur_led->gpio, cur_led->name);
+ if (ret < 0)
+ goto err;
+
led_dat->cdev.name = cur_led->name;
led_dat->cdev.default_trigger = cur_led->default_trigger;
led_dat->gpio = cur_led->gpio;
@@ -87,10 +91,6 @@ static int gpio_led_probe(struct platform_device *pdev)
led_dat->cdev.brightness_set = gpio_led_set;
led_dat->cdev.brightness = LED_OFF;
- ret = gpio_request(led_dat->gpio, led_dat->cdev.name);
- if (ret < 0)
- goto err;
-
gpio_direction_output(led_dat->gpio, led_dat->active_low);
INIT_WORK(&led_dat->work, gpio_led_work);
--
1.5.4.3