samples: threads: update to use new GPIO API #21116
Conversation
Update sample application to use new GPIO API: - GPIO flags defined by the devicetree - replace gpio_pin_write with gpio_pin_set function Signed-off-by: Piotr Mienkowski <piotr.mienkowski@gmail.com>
| __ASSERT_NO_MSG(gpio_dev != NULL); | ||
| gpio_dev = device_get_binding(led->gpio_dev_name); | ||
| __ASSERT(gpio_dev != NULL, "Error: didn't find %s device\n", | ||
| led->gpio_dev_name); |
There was a problem hiding this comment.
Would be more consistent sing printk, similarly as what is done below
There was a problem hiding this comment.
The blinky and button samples use printk in this place so this is indeed not consistent. However, the ASSERT version is closer to what the real application would look like. Checking the result of device_get_binding is only required during the development phase. Once we provide a correct device name the check can be turned off. On the other hand the call to gpio_pin_configure can fail during the run time if we are working with an external GPIO driver.
I would rather stay with the existing approach.
There was a problem hiding this comment.
The blinky and button samples use printk in this place so this is indeed not consistent. However, the ASSERT version is closer to what the real application would look like. Checking the result of device_get_binding is only required during the development phase. Once we provide a correct device name the check can be turned off.
This makes sense, but from what I see this isn't a widely used approach accros samples.
I'm approving the change, but it would be interesting to have this proposed (and potentially discussed) as a general guidelines for samples implementation.
| __ASSERT_NO_MSG(gpio_dev != NULL); | ||
| gpio_dev = device_get_binding(led->gpio_dev_name); | ||
| __ASSERT(gpio_dev != NULL, "Error: didn't find %s device\n", | ||
| led->gpio_dev_name); |
There was a problem hiding this comment.
The blinky and button samples use printk in this place so this is indeed not consistent. However, the ASSERT version is closer to what the real application would look like. Checking the result of device_get_binding is only required during the development phase. Once we provide a correct device name the check can be turned off.
This makes sense, but from what I see this isn't a widely used approach accros samples.
I'm approving the change, but it would be interesting to have this proposed (and potentially discussed) as a general guidelines for samples implementation.
Update sample application to use new GPIO API: