Closed GeVanCo closed 1 year ago
Having the blank layer at the end of installation is the correct approach. Activation of the rpi driver should be done after install on the “Hardware” tab. There are several rpi based boards (like UniPi for example) that will just break if the rpi driver is installed automatically
Ok, I was missing that part of the build flow. Will revert the changes in the script in a next commit
Reverted back the changes I did in background_installer.sh
since I didn't know the target to compile against was chosen the moment you select the hardware layer in the web client.
So, this commit only contains the necessary changes in the raspberrypi.cpp
file to work correctly.
Currently, the RPi HW layer file is not compiled at all since the
background_installer
script was invoking thechange_hardware_layer
script incorrectly. The parameterblank_script
was passed instead of the parameterrpi
. Due to this change, the fileraspberrypi.cpp
is now compiled (was not in the past, it was just the empty blank file) and installed when running the compilation script./install.sh
with the argumentrpi
.Next to this, the content of the file
raspberrypi.cpp
was wrong. When iterating over the arrays, the index parameteri
of thefor
-loop was passed to the functionpinNotPresent()
instead of the item at locationi
of the corresponding array. This has been corrected, is tested and works perfect now.