nRF5340-DK network core reboots running hci_rpmsg when configured with CONFIG_SERIAL=n

[d_c_h_w]

I am trying to do some power measurements on the nRF5340 using the power profiler.

As suggested in the guide here, I am removing serial support to reduce power consumption, which I need to do on both cores

https://developer.nordicsemi.com/nRF_Connect_SDK/doc/1.5.0/nrf/app_power_opt.html

When I tested this a while ago on a PDK with an earlier SDK it worked, but since upgrading to SDK 1.5.0 and using a DK board , I see the network core (running hci_rpmsg) constantly reboot.

I just get a long string of the following coming out of the virtual com port (if I do connect to it)

*** Booting Zephyr OS build v2.4.99-ncs1  ***

*** Booting Zephyr OS build v2.4.99-ncs1  ***

*** Booting Zephyr OS build v2.4.99-ncs1  ***

*** Booting Zephyr OS build v2.4.99-ncs1  ***

The app core meanwhile is waiting for a hci related semaphore, gives up after a while and asserts.

If I (painfully) roll everything back to v1.4.2 the net core is happy again. I get the following and it continues to operate, although as serial is disabled I get nothing further out of the virtual com port.

*** Booting Zephyr OS build v2.4.0-ncs2 ***
Booting (0x1008800).

So it looks like this use case has got broken somewhere along the line. Anyone else seeing this?

[Carl Richard]

Hi, d_c_h_w!

Try to add CONFIG_LOG=n to the hci_rpmsg configuration. The logging interface depends on serial by default, but is unfortunately not disabled when serial is. 

Please reach out if you face any other issues.

Best regards,
Carl Richard