debugging a coredump with gdb does not show full backtrace

[ziv123]

hello Nordic

i am working with nrf 52840 and 52832. using ncs v2.8.0

i am saving a coredump to flash partition and i am using gdb server - client to try to parse the coredump .bin file (as shown here: https://docs.nordicsemi.com/bundle/ncs-2.8.0/page/zephyr/services/debugging/coredump.html#:~:text=Run%20the%20core%20dump%20serial%20log%20converter%3A)

(i am not using coredump_serial_log_parser.py since i already have a .bin file in flash) in gdb when i run 'bt full' i only get the place of crash and not a full backtrace ..

in my device i build with  CONFIG_DEBUG_COREDUMP_MEMORY_DUMP_THREADS=y  (as far as i know it should be enough)

any idea what i am missing  ?

hope to read you soon

best regards

Ziv

[Håkon Alseth]

Hi,

 

I would recommend that you contact the zephyr community related to coredump functionality.

Specifically due to issue(s) like this:

https://github.com/zephyrproject-rtos/zephyr/issues/78788

 

Ie. it is then important to use a newer ncs version (>=v3.0.0).

 

Kind regards,

Håkon

[ziv123]

currently i am working with ncs v2.8.0 which works with zephyr v3.7.99-ncs1,

ncs 2.9.2. is working with zephyr v3.7.99-ncs2

how can i find which zephyr version is on under a certain ncs .. lets say ncs v3.0 or ncs 3.1 ?

why where is no release for ncs v3.0. and there seems to be a jump from ncs v2.9.2 to ncs v3.1. ?

best regards

Ziv

[Håkon Alseth]

Hi,

 

You can follow the nrf/west.yml file for the release in question:

https://github.com/nrfconnect/sdk-nrf/blob/v3.0.2/west.yml#L68

ziv123 said:

why where is no release for ncs v3.0. and there seems to be a jump from ncs v2.9.2 to ncs v3.1. ?

NCS v3.0.x is present.

 

Kind regards,

Håkon

[ziv123]

hi 

well indeed after migrating to ncs3.0.2 it seems to have full back trace

however, can you explain why this module "ncs/nrf/subsys/debug/coredump/coredump_backend_nrf_flash_partition.c was removed after ncs2.8.0 ? is there support for coredump collection for nrf devices on another place ? Also, why are there no .py scripts for dealing with binary files of a coredump downloaded from device's flash but just one to handle logs of coredump ? (i fixed something by myself but still seems like missing) 

hope to read you soon

best regards

Ziv

[Håkon Alseth]

Hi,

 

ziv123 said:

however, can you explain why this module "ncs/nrf/subsys/debug/coredump/coredump_backend_nrf_flash_partition.c was removed after ncs2.8.0 ? is there support for coredump collection for nrf devices on another place ?

I am not sure what you mean, the file is present:

https://github.com/nrfconnect/sdk-nrf/tree/main/subsys/debug/coredump

ziv123 said:

Also, why are there no .py scripts for dealing with binary files of a coredump downloaded from device's flash but just one to handle logs of coredump ? (i fixed something by myself but still seems like missing) 

It seems that, based on the core dump documentation, zephyr does not provide means to read out the content over a given transport:

https://docs.zephyrproject.org/latest/services/debugging/coredump.html

You can use nrfjprog or nrfutil, or jlink tools, to readback the memory content to for instance a .bin.

 

Kind regards,

Håkon