Hi @JerryChang,
I’ve now managed to fix my image regarding the /sys/kernel/debug path.
It seems I was getting a bug similar to that in corrupted-kernel-headers-package-in-l4t-32-6-1. I downloaded the BSP through the download page instead of the SDK and this error seems to have been removed so I now have access to files in the /sys/kernel/debug folder
It turns out that the setting of the default value was actually working and I was seeing similar to this post jetson-nano-gpio-example-problem
When setting the pins to default high, the cheap DMM I was using was loading the GPIO pins pulling them low. This was only verified by scoping the pins during the carrier board reset which showed that they begun high and when probed with DMM were driven low. It was even the case that insertion of the scope probes (x10) at times was possible to drive the pins low!
This factor is unfortunate as it doesn’t help with our application as we require the pins held low. It seems like a bit of capacitive coupling can actually cause the pins to float.
When I use the Jetson.GPIO library however it seems like by setting the GPIO via sysfs the pins remain high/low when probed with the DMM. I would like that similar drive strength when setting the default pin configuration.
I’m not sure that it’s possible but have opened a separate thread to discuss that possibility here: overcoming-gpio-capacitive-loading-at-startup