nabaztag2018 / pynab

Nabaztag en Python pour RaspberryPi
GNU General Public License v3.0
161 stars 53 forks source link

Regression : raspberry pi zero yellow led is blinking #256

Closed SonnyT72 closed 3 years ago

SonnyT72 commented 3 years ago

Version of pynab: v0.9.0 (updated from previous available release)

Version of Nabaztag:

Version of TagTagTag:

The problem Regression : the raspberry pi zero internal yellow led, disabled in a previous release of pynab, is back to life on my rabbit following 0.9.0 pynab update. Update was succesfully done through the web interface.

Expected behavior The led shall be shut off after startup.

f-laurens commented 3 years ago

What exact version of the Raspberry Pi kernel are you running? (see Version information on the System Information page).

SonnyT72 commented 3 years ago

I am running Raspbian GNU/Linux 10 (buster) - Kernel 4.19.118+ armv6l

f-laurens commented 3 years ago

Kernel 4.19.118+ armv6l

This is not really a Pynab bug, but rather a Raspberry Pi kernel related glitch: the logic to toggle the Pi Zero LED on/off was inverted in kernel 5.x :-( A solution for you is to update your kernel from 4.x to 5.x:

SonnyT72 commented 3 years ago

Thank you for your answer! I will try the first solution now and give the result here for others to know. If I don't succeed I will then switch to your second solution.

Thanks again to you and the team working hard to bring our rabbits back to life!

SonnyT72 commented 3 years ago

If I find how to remove it, I will :)

mehdichaouch commented 3 years ago

In the meanwhile, can you remove the bug label from this issue?

As I know labels can be managed only by @nabaztag2018

pguyot commented 3 years ago

I beg to differ: this is a bug of pynab. We released 4.x images and allowed @SonnyT72 and others to upgrade our code without upgrading the kernel. We definitely should check kernel version and handle the led accordingly.

f-laurens commented 3 years ago

We definitely should check kernel version and handle the led accordingly.

Indeed; PR #158 for Issue #156 should not have been merged as is. PR #257 does it better...