Rock64 Armbian

[chazhenry]

I've git cloned, but can't get this working under Armbian for Rock64. The /sys/class/gpio/export file is present, but not accessible. The gpio101 and gpio102 files are not present. Stuck on this for days. Tried Ubuntu, but it's unstable. Get constant seg faults.

`root@rock64:/home/sas# python3 gpio_test.py
Testing R64.GPIO Module...

Module Variables:
Name Value

---

GPIO.ROCK ROCK
GPIO.BOARD BOARD
GPIO.BCM BCM
GPIO.OUT out
GPIO.IN in
GPIO.HIGH 1
GPIO.LOW 0
GPIO.PUD_UP 0
GPIO.PUD_DOWN 1
GPIO.VERSION 0.6.3
GPIO.RPI_INFO {'RAM': '1024M', 'PROCESSOR': 'BCM2837', 'TYPE': 'Pi 3 Model B', 'REVISION': 'a22082', 'P1_REVISION': 3, 'MANUFACTURER': 'Embest'}
Error: Unable to export GPIO
Error: Unable to set GPIO direction
Error: Unable to export GPIO
Error: Unable to set GPIO direction

Testing GPIO Input/Output:
You must setup() the GPIO channel ({0} {1}) first
Traceback (most recent call last):
File "gpio_test.py", line 42, in
var_gpio_state = GPIO.input(var_gpio_out) # Return state of GPIO
File "/home/sas/R64/_GPIO.py", line 210, in input
if validate_direction(channel, 'both') == 0:
File "/home/sas/R64/_GPIO.py", line 118, in validate_direction
print("You must setup() the GPIO channel ({0} {1}) first").format(gpio_mode, channel)
AttributeError: 'NoneType' object has no attribute 'format'`

[ghost]

There are a couple of typos in _GPIO.py when using the format function
print("You must setup() the GPIO channel ({0} {1}) first").format(gpio_mode, channel)
and others have to be changed to
print("You must setup() the GPIO channel ({0} {1}) first".format(gpio_mode, channel))

[Leapo]

Those aren't typos... it was written that way intentionally, and runs just fine. I even flashed the latest version of Armbian on my Rock64 to test, and had no issues with my current code.

In fact, making your suggested changes causes the library to become completely non-functional on my Rock64. Can you explain why this might be happening?

Edit: Also, chazhenry's post shows that the command runs fine, without modification. It's printing the expected debug text to screen.

[ghost]

It's printing the expected debug text to screen.

Actually I don't think so, it is a traceback message. See the last line, it states it is an attribute error.
The NoneType is the result of the print function. The code is, as far as I'm aware, only valid
python2 code if standard print statement is used but it throws this execption if python3, or python3 like print function, would be used.

[Leapo]

I'll note, even if the suggested code needs to be changed, it doesn't solve his problem. Python is only trying to execute the debug text because no GPIO is set up. We need to figure out why GPIO setup is failing to actually solve chazhenry's problem.

@chazhenry Can you try to export the GPIO manually from Bash and see if it works?

[Leapo]

@ClaudiaFrank After testing with both the Python2 and Python3 console, you're right about the debug text failing under Python3. I'll work on that.

Edit: Updated master with fixed print statements. Thanks for the suggestion!

[ghost]

@Leapo, at the moment, I don't have access to a Rock64, the one I was testing is in use by a friend of mine.
I will ask him if he still needs it or whether I can borrow it to make some tests. I will let you know.

[ghost]

Sorry, I received the rock64 but it seems broken, so I can't be any help in testing.

[limpid-kzonix]

I can say that everything works fine on Rock64. I had checked GPIO Model on ROCK64 Ver3 in my nodejs project by using basic-gpio package...