We are developing a mesh application for nRF5340 currently using the nRF5340PDK. We have loaded one dev board with the Mesh Light example and the other with the Mesh Light Switch example (including adding the following to prj.conf: CONFIG_BT_CTLR_TX_BUFFER_SIZE=74, CONFIG_BT_CTLR_DATA_LENGTH_MAX=74). Provisioning is succesfull for both boards and the Mesh Light is controllable through the nRF Mesh app. However after having setup the Mesh Light Switch to publish to all address (0xFFFF) the relavent LED on the Light Switch illuminates but not on the Light when the Light Switch is pressed. After some fiddling with publish settings on th eLight Switch it appears that increasing the retransmission count to 7 does yield some control of the Mesh Light however only ~50% of the time. Please advise if this is expected or if there is a way to fix this. We are reluctant to press on with custom mesh application development while the example is not working reliably. Thanks in advance.
In addition to this issue we face Zephyr stack overflow tracing back to the log_backend_uart.c thread when enabling mesh logging for ACCESS, TRANS, NET, ADV. This is making it very difficult to diagnose the above issue.
# Logging CONFIG_BT_DEBUG_LOG=y CONFIG_BT_MESH_DEBUG=y CONFIG_BT_MESH_DEBUG_MODEL=y CONFIG_BT_MESH_DEBUG_PROV=y CONFIG_BT_MESH_DEBUG_BEACON=y CONFIG_BT_MESH_DEBUG_ACCESS=y CONFIG_BT_MESH_DEBUG_TRANS=n CONFIG_BT_MESH_DEBUG_NET=n CONFIG_BT_MESH_DEBUG_ADV=n
[00:00:29.450,592] <err> os: ***** USAGE FAULT ***** [00:00:29.450,592] <err> os: Stack overflow (context area not valid) [00:00:29.450,592] <err> os: r0/a1: 0x0001f0f2 r1/a2: 0x21000000 r2/a3: 0x01000000 [00:00:29.450,622] <err> os: r3/a4: 0x02000000 r12/ip: 0x04000000 r14/lr: 0x08000000 [00:00:29.450,622] <err> os: xpsr: 0x20000000 [00:00:29.450,622] <err> os: Faulting instruction address (r15/pc): 0x10000000 [00:00:29.450,653] <err> os: >>> ZEPHYR FATyr OS build v2.3.0-rc1-ncs1 ***
