SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

RE: SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

Petter Myhre — Wed, 24 Aug 2016 08:39:49 GMT

Thank you for the update.

RE: SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

Eliot Stock — Fri, 19 Aug 2016 19:37:10 GMT

This was down to an error in the linker script for the application. We use the example bootloader from SDK 11 and I had wrongly assumed that the .fs_data SECTION in the application's linker script should match the one in the bootloader. It should not. The application uses the fstorage module, the bootloader doesn't. Best explained here:

devzone.nordicsemi.com/.../

RE: SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

Eliot Stock — Thu, 18 Aug 2016 14:42:14 GMT

Had to open a case, sorry. Thanks.

RE: SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

Petter Myhre — Thu, 18 Aug 2016 13:58:35 GMT

Sounds like a plan.

RE: SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

Eliot Stock — Thu, 18 Aug 2016 13:50:56 GMT

Sure. I'll just step back through some old checkins and see if a change I made has caused it. If I can't pin it down I'll open a case with the project attached - too much code to post here. Thanks.

RE: SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

Petter Myhre — Thu, 18 Aug 2016 13:37:33 GMT

I guess the error is coming from somewhere else then. I would like to help you to debug this, could you share your project? Or a project that reproduces the issue?

RE: SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

Eliot Stock — Thu, 18 Aug 2016 12:55:07 GMT

Line numbers above are for security_dispatcher.c, not peer_id.c, btw.

RE: SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

Eliot Stock — Thu, 18 Aug 2016 12:54:17 GMT

That's exactly what's confusing me - that line with err_code = NRF_ERROR_INTERNAL, L588 does NOT execute. err_code is still 0 at L592. Nothing in the block that follows is getting a non-zero err_code when I debug it. No error at L592, nor L596, nor L612. And yet at L621, err_code is 3. (These are all breakpoints set in isolation and I'm reflashing and resetting before testing each one, I'm not expecting to step through code while in a BLE connection, just to be clear.)

RE: SDK 11 peer manager: smd_params_reply() returns NRF_ERROR_INTERNAL

Petter Myhre — Thu, 18 Aug 2016 12:15:14 GMT

If you put a breakpoint at err_code = NRF_ERROR_INTERNAL;, do you hit it? If you do, where in PM_PEER_ID_INVALID returned from inside pdb_peer_allocate()?