make ledtrig-gpio useable with GPIO drivers requiring threaded irqs
These patches make it possible to use the ledtrig-gpio driver with GPIO drivers that require threaded IRQs (like the PCA953x I2C driver). It has been tested with a PCA9554 chip on an i.MX28 module.
PATCH 1/2 leds: trigger: gpio: fix warning in gpio trigger for gpios whose accessor function may sleep
When using a GPIO driver whose accessor functions may sleep (e.g. an I2C GPIO extender like PCA9554) the following warning is issued:
WARNING: CPU: 0 PID: 665 at drivers/gpio/gpiolib.c:2274 gpiod_get_raw_value+0x3c/0x48() Modules linked in: CPU: 0 PID: 665 Comm: kworker/0:2 Not tainted 3.16.0-karo+ #115 Workqueue: events gpio_trig_work [<c00142cc>] (unwind_backtrace) from [<c00118f8>] (show_stack+0x10/0x14) [<c00118f8>] (show_stack) from [<c001bf10>] (warn_slowpath_common+0x64/0x84) [<c001bf10>] (warn_slowpath_common) from [<c001bf4c>] (warn_slowpath_null+0x1c/0x24) [<c001bf4c>] (warn_slowpath_null) from [<c020a1b8>] (gpiod_get_raw_value+0x3c/0x48) [<c020a1b8>] (gpiod_get_raw_value) from [<c02f68a0>] (gpio_trig_work+0x1c/0xb0) [<c02f68a0>] (gpio_trig_work) from [<c0030c1c>] (process_one_work+0x144/0x38c) [<c0030c1c>] (process_one_work) from [<c0030ef8>] (worker_thread+0x60/0x5cc) [<c0030ef8>] (worker_thread) from [<c0036dd4>] (kthread+0xb4/0xd0) [<c0036dd4>] (kthread) from [<c000f0f0>] (ret_from_fork+0x14/0x24) ---[ end trace cd51a1dad8b86c9c ]---
Fix this by using the _cansleep() variant of gpio_get_value().
--- drivers/leds/trigger/ledtrig-gpio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/leds/trigger/ledtrig-gpio.c +1 -1
@@ -48,7 +48,7 @@ static void gpio_trig_work(struct work_struct *work)if (!gpio_data->gpio)return ; - tmp =gpio_get_value (gpio_data->gpio); + tmp =gpio_get_value_cansleep (gpio_data->gpio);if (gpio_data->inverted) tmp = !tmp;
Signature · 1 line
1.7.10.4
Re: [PATCH 1/2] leds: trigger: gpio: fix warning in gpio trigger for gpios whose accessor function may sleep
Quoted 17 lines · Lothar Waßmann
When using a GPIO driver whose accessor functions may sleep (e.g. an I2C GPIO extender like PCA9554) the following warning is issued:
WARNING: CPU: 0 PID: 665 at drivers/gpio/gpiolib.c:2274 gpiod_get_raw_value+0x3c/0x48() Modules linked in: CPU: 0 PID: 665 Comm: kworker/0:2 Not tainted 3.16.0-karo+ #115 Workqueue: events gpio_trig_work [<c00142cc>] (unwind_backtrace) from [<c00118f8>] (show_stack+0x10/0x14) [<c00118f8>] (show_stack) from [<c001bf10>] (warn_slowpath_common+0x64/0x84) [<c001bf10>] (warn_slowpath_common) from [<c001bf4c>] (warn_slowpath_null+0x1c/0x24) [<c001bf4c>] (warn_slowpath_null) from [<c020a1b8>] (gpiod_get_raw_value+0x3c/0x48) [<c020a1b8>] (gpiod_get_raw_value) from [<c02f68a0>] (gpio_trig_work+0x1c/0xb0) [<c02f68a0>] (gpio_trig_work) from [<c0030c1c>] (process_one_work+0x144/0x38c) [<c0030c1c>] (process_one_work) from [<c0030ef8>] (worker_thread+0x60/0x5cc) [<c0030ef8>] (worker_thread) from [<c0036dd4>] (kthread+0xb4/0xd0) [<c0036dd4>] (kthread) from [<c000f0f0>] (ret_from_fork+0x14/0x24) ---[ end trace cd51a1dad8b86c9c ]---
Quoted 2 lines
Fix this by using the
_cansleep()variant ofgpio_get_value().
Good catch, I will merge this.
Thanks,
-Bryan
Quoted 21 lines
Signed-off-by Lothar Waßmann <[email protected]>--- drivers/leds/trigger/ledtrig-gpio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/leds/trigger/ledtrig-gpio.c +1
-2@@ -48,7 +48,7 @@ static void gpio_trig_work(struct work_struct *work)if (!gpio_data->gpio)return ; - tmp =gpio_get_value (gpio_data->gpio); + tmp =gpio_get_value_cansleep (gpio_data->gpio);if (gpio_data->inverted) tmp = !tmp; --1.7.10.4
PATCH 2/2 leds: trigger: gpio: make ledtrig-gpio useable with GPIO drivers requiring threaded irqs
In reply to Lothar WaßmannWhen trying to use the LED GPIO trigger with e.g. the PCA953x GPIO driver, request_irq() fails with -EINVAL, because the GPIO driver requires a nested interrupt handler.
Use request_any_context_irq() to be able to use any GPIO driver as LED trigger.
--- drivers/leds/trigger/ledtrig-gpio.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/leds/trigger/ledtrig-gpio.c +3 -3
@@ -161,10 +161,10 @@ static ssize_t gpio_trig_gpio_store(struct device *dev,return n; } - ret =request_irq (gpio_to_irq (gpio), gpio_trig_irq, + ret =request_any_context_irq (gpio_to_irq (gpio), gpio_trig_irq,IRQF_SHARED |IRQF_TRIGGER_RISING |IRQF_TRIGGER_FALLING ,"ledtrig-gpio" , led); -if (ret) { +if (ret <0 ) {dev_err (dev,"request_irq failed with error %d\n" , ret); }else {if (gpio_data->gpio !=0 )
@@ -172,7 +172,7 @@ static ssize_t gpio_trig_gpio_store(struct device *dev, gpio_data->gpio = gpio; } -return ret ? ret : n; +return ret <0 ? ret : n; } static DEVICE_ATTR(gpio, 0644, gpio_trig_gpio_show, gpio_trig_gpio_store);
Signature · 1 line
1.7.10.4
Re: [PATCH 2/2] leds: trigger: gpio: make ledtrig-gpio useable with GPIO drivers requiring threaded irqs
Quoted 7 lines · Lothar Waßmann
When trying to use the LED GPIO trigger with e.g. the PCA953x GPIO driver,
request_irq()fails with-EINVAL, because the GPIO driver requires a nested interrupt handler.Use
request_any_context_irq()to be able to use any GPIO driver as LED trigger.
Hmmm, what about use request_thread_irq() and put the gpio_trig_work() in as the thread_func.
Felipe, can you take a look at this?
Also in the first patch:
Actually in gpio_trig_irq(), it said:
/* just schedule_work since gpio_get_value can sleep */
schedule_work(&gpio_data->work);
Then that means we need to call gpio_get_value_can_sleep() in the gpio_trig_work() instead of gpio_get_value(), right?
Thanks,
-Bryan
Quoted 34 lines
Signed-off-by Lothar Waßmann <[email protected]>--- drivers/leds/trigger/ledtrig-gpio.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)drivers/leds/trigger/ledtrig-gpio.c +3
-4@@ -161,10 +161,10 @@ static ssize_t gpio_trig_gpio_store(struct device *dev,return n; } - ret =request_irq (gpio_to_irq (gpio), gpio_trig_irq, + ret =request_any_context_irq (gpio_to_irq (gpio), gpio_trig_irq,IRQF_SHARED |IRQF_TRIGGER_RISING |IRQF_TRIGGER_FALLING ,"ledtrig-gpio" , led); -if (ret) { +if (ret <0 ) {dev_err (dev,"request_irq failed with error %d\n" , ret); }else {if (gpio_data->gpio !=0 )@@ -172,7 +172,7 @@ static ssize_t gpio_trig_gpio_store(struct device *dev, gpio_data->gpio = gpio; } -return ret ? ret : n; +return ret <0 ? ret : n; } static DEVICE_ATTR(gpio, 0644, gpio_trig_gpio_show, gpio_trig_gpio_store); --1.7.10.4
Re: [PATCH 2/2] leds: trigger: gpio: make ledtrig-gpio useable with GPIO drivers requiring threaded irqs
Hi,
Quoted 14 lines · Bryan Wu
Quoted 7 lines · Lothar Waßmann
When trying to use the LED GPIO trigger with e.g. the PCA953x GPIO driver,
request_irq()fails with-EINVAL, because the GPIO driver requires a nested interrupt handler.Use
request_any_context_irq()to be able to use any GPIO driver as LED trigger.Hmmm, what about use
request_thread_irq()and put thegpio_trig_work()in as thethread_func.Felipe, can you take a look at this?
Quoted 8 lines
Also in the first patch:
Actually ingpio_trig_irq(), it said:
/* justschedule_worksincegpio_get_valuecan sleep */schedule_work(&gpio_data->work);Then that means we need to call
gpio_get_value_can_sleep()in thegpio_trig_work()instead ofgpio_get_value(), right?
That's exactly what my first patch does!
Lothar Waßmann
Signature · 9 lines
___________________________________________________________ Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Geschäftsführer: Matthias Kaussen Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 www.karo-electronics.de | [email protected] ___________________________________________________________
Re: [PATCH 2/2] leds: trigger: gpio: make ledtrig-gpio useable with GPIO drivers requiring threaded irqs
Quoted 22 lines · Lothar Waßmann
Hi,
Quoted 19 lines · Bryan Wu
Quoted 7 lines · Lothar Waßmann
When trying to use the LED GPIO trigger with e.g. the PCA953x GPIO driver,
request_irq()fails with-EINVAL, because the GPIO driver requires a nested interrupt handler.Use
request_any_context_irq()to be able to use any GPIO driver as LED trigger.Hmmm, what about use
request_thread_irq()and put thegpio_trig_work()in as thethread_func.Felipe, can you take a look at this?
Also in the first patch:
Actually ingpio_trig_irq(), it said:
/* justschedule_worksincegpio_get_valuecan sleep */schedule_work(&gpio_data->work);
Quoted 5 lines
Quoted 3 lines
Then that means we need to call
gpio_get_value_can_sleep()in thegpio_trig_work()instead ofgpio_get_value(), right?That's exactly what my first patch does!
Yeah, exactly. I'm just curious about the comment here.
-Bryan
Quoted 12 lines
Lothar Waßmann
--
___________________________________________________________Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen
Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Geschäftsführer: Matthias Kaussen
Handelsregistereintrag: Amtsgericht Aachen, HRB 4996www.karo-electronics.de | [email protected]
___________________________________________________________