pm_init hides real cause of error

[goldwake]

Hi Nordic,

I have recently received an error from pm_init that I have never seen before, NRF_ERROR_INTERNAL. After some debugging, I found it was really due to NRF_ERROR_STORAGE_FULL (I'm not sure how it got to this state).

Why does pm_init hide the true cause of the error?

As a side note: Ideally my code should handle the freeing of memory but, seems a bit silly that if I got the error due to NRF_ERROR_STORAGE_FULL, that I can't use pm_peer_delete or pm_peers_delete to fix the problem because the peer manager is not initialised. Is there a way to free this data without deleting my own data stored using FDS?

I'm using SDK13.0

[Petter Myhre]

I agree that more errors should be propagated to pm_init(), but atm. NRF_ERROR_INTERNAL is a collection of errors, and you need to go into the internals to figure out where the error comes from. Exactly which function is returning the error? Please also include the hexadecimal value of the error. Is the flash full? Have you bonded with many (or any) devices?

[ndarkness]

Any answer to this issue? I am facing the same issue

[goldwake]

Not really. Other than trying to prevent it happening in the first place, if we get that error we wipe the flash and try again

[????]

after I use fds for data storage，the same issue with me when I try to ota， the first hex works fine but can not work after ota any version