Zephyr: STM32wb55 bluetooth samples fail

[yhvh]

Using a early version of the STM32wb55 most of the Bluetooth samples fail to connect with either Android app or bluetoothctl. As mentioned in

14188 this is possibly because of a need to upgrade the M0 core to one which works with zephyr. How do I go about doing that?

Ping @erwango

Steps to reproduce the behavior:

1. Build one of samples/Bluetooth/peripheral*
2. Flash to early version STM32wb55
3. Attempt to connect via bluetoothctl
4. Observe lack of connection

Expected behavior
The connection should succeed

Impact
Unable to use #14188

Environment (please complete the following information):

• OS: Linux
• Toolchain: Zephyr SDK
• #14188

Additional context
STM32wb55 early release version

[carlescufi]

I understand from @erwango that the early versions of the devkit are not supported in Zephyr

[erwango]

Reopening
Early sample versions are indeed not supported for now, but it makes sense to check what it costs to get them working.

@yhvh, as a start: can you try to get a STM32WB stm32cube package from st.com and follow the steps indicated to update M0 f/w (in Projects/STM32WB_Copro_Wireless_Binaries/Release_Notes.html, using stm32wb5x_BLE_Stack_fw.bin)?

[yhvh]

@erwango This worked, following the instructions for the f/w version on the board

• Upgraded FUS, then the BLE stack.
• Tested the CRC and HR peripherals on top of #14188 with nRF android application
• Observed notification stream

Thanks!

[erwango]

Closing (and keeping as reference for future similar question).

[pavel-demin]

I have the same problem with P-NUCLEO-WB55 EPNWB55$CU4. STM32WB55 on this board doesn't have FUS and newer versions of the M0 encrypted firmware can't be installed on this board. Would it be possible to make this kind of board working with Zephyr's BLE host?

The problem is the same with the peripheral* and st_ble_sensor samples. The host stops communicating with the controller right after connection, possibly due to an unhandled event:

[00:00:06.586,000] <inf> hci_ipm: EVT: hcievt->evtserial.evt.evtcode: 0x3e
[00:00:06.586,000] <inf> main: Connected
[00:00:06.587,000] <inf> hci_ipm: CMD: code 0x2016 plen 2
[00:00:06.587,000] <inf> hci_ipm: EVT: hcievt->evtserial.evt.evtcode: 0x0f
[00:00:06.782,000] <inf> hci_ipm: EVT: hcievt->evtserial.evt.evtcode: 0x01
[00:00:06.782,000] <wrn> bt_hci_core: Unhandled event 0x01 len 40: 07000300040002b9000000000000000~

[00:00:36.816,000] <inf> hci_ipm: EVT: hcievt->evtserial.evt.evtcode: 0x05
[00:00:36.817,000] <inf> main: Disconnected (reason 19)
[00:00:36.817,000] <inf> hci_ipm: CMD: code 0x200a plen 1
[00:00:36.817,000] <inf> hci_ipm: EVT: hcievt->evtserial.evt.evtcode: 0x0e

It seems that something is wrong with the HCI_COMMAND_STATUS_EVENT response to the HCI_LE_READ_REMOTE_FEATURES command.

[erwango]

@pavel-demin I'm truly sorry but it is not possible to provide support on these samples since the binary is quite specific.