Here are two test images for RPi4 and Generic (GBM) that contain connman/iwd with latest patches/changes (HEAD):
https://chewitt.libreelec.tv/testing/LibreE….arm-11.0.1.tar
https://chewitt.libreelec.tv/testing/LibreE…6_64-11.0.1.tar
If others can confirm things are resolved we'll get the changes submitted and merged.
Thanx for the very fast build;
some autoconnects, some not, I grabbed that scenario: failure in GUI (no autoconnect), but was able to manually connect by entering the passphrase - so no invalid key - then connect - around 5 boots, 2x passphrase requests, 3 okay.
Your log chewitt53-test-no-autoconnect.log -> http://ix.io/4rTi
I guess you compiled iwd master like me? But is it wise to change more than one software part when tracking a bug (I mean connman here). Just my thought... I only replaced iwd and checked that with different LE versions.
There is some difference in the iwds
# ldd /usr/lib/iwd <- your version #53
/usr/lib/libarmmem-v7l.so => /usr/lib/libarmmem-v7l.so (0xf7ce0000)
libc.so.6 => /usr/lib/libc.so.6 (0xf7b70000)
/lib/ld-linux-armhf.so.3 => /usr/lib/ld-linux-armhf.so.3 (0xf7d09000)
ldd /storage/bin/iwd <- on OSMC compiled version
/usr/lib/libarmmem-v7l.so => /usr/lib/libarmmem-v7l.so (0xf7d00000)
libdl.so.2 => /usr/lib/libdl.so.2 (0xf7cd0000)
libc.so.6 => /usr/lib/libc.so.6 (0xf7b60000)
/lib/ld-linux-armhf.so.3 => /usr/lib/ld-linux-armhf.so.3 (0xf7d30000)
Could be this relevant? Your LE version doesn't use libdl.so.2