nrf9160 reboots sporadically after some time.

RE: nrf9160 reboots sporadically after some time.

J0sh1101 — Mon, 03 Apr 2023 07:33:14 GMT

Hi,

after careful debugging and going through the code, I found that I had memory leak with the cJson library. I did not free all the memory. I freed the the cJson object but not the char* memory buffer.

Didrik Rokhaug Thanks for the help

This is how it should look when you send an mqtt message using the cJson library:

RE: nrf9160 reboots sporadically after some time.

Didrik Rokhaug — Fri, 24 Mar 2023 15:26:28 GMT

[quote user="J0sh1101"]Should I add the CONFIG_TFM_SECURE_UART=y?[/quote]

Yes.

[quote user="J0sh1101"]I noticed that the error occurs whenever there is a context switch in the arch_swap function. [/quote]

Do you have a stack trace before it enters arch_swap?

[quote user="J0sh1101"]hen I try to remove them, I get an error message. (I have used the AWS IOT example as a base).[/quote]

What error message do you get?

The options shouldn't cause the crash you are seeing, but I also cannot see why they cause an error when you remove them.

RE: nrf9160 reboots sporadically after some time.

J0sh1101 — Thu, 23 Mar 2023 20:33:05 GMT

Hello,

I have cleaned up my code and updated prj.conf, but I still end up in the Secure Hard Fault Handler. Now, every time I try to connect to the cloud, using the mqtt_conect(), I am landing in the secure hard fault handler. I tried increasing the mqtt buffers, and also increased the heap and stack to 4 kB, but it didn't change anything. It's a bit annoying that I can't look back in the stack frame to see where the error is occurring.

I noticed that the error occurs whenever there is a context switch in the arch_swap function. There is probably a backend in Zephyr that does all the work with sockets and that is where the exception is thrown.

Are there any other options in zephyr or nordic where I can change something in the pry.conf file to get more logs or information? Or call a function to give me more info?

Thanks for your time and help!

RE: nrf9160 reboots sporadically after some time.

J0sh1101 — Thu, 23 Mar 2023 09:40:24 GMT

I added CONFIG_TFM_SECURE_UART_SHARE_INSTANCE to prj.conf, but I did not notice any difference. After hovering over the CONFIG_TFM_SECURE_UART_SHARE_INSTANCE , I got this text:

CONFIG_TFM_SECURE_UART_SHARE_INSTANCE was assigned the value y, but got the value n. Missing dependencies:
TFM_SECURE_UART

Should I add the CONFIG_TFM_SECURE_UART=y?

In proj.conf I have the AWS_IOT flags in there, even though I don't use the aws_iot library. Is this ok? Could this produce the problem. When I try to remove them, I get an error message. (I have used the AWS IOT example as a base).

I have this flags:

CONFIG_AWS_IOT=y

CONFIG_AWS_IOT_CLIENT_ID_STATIC="aero_sens"

CONFIG_AWS_IOT_BROKER_HOST_NAME="example.endpoint.com"

CONFIG_AWS_IOT_SEC_TAG=201

CONFIG_AWS_IOT_APP_SUBSCRIPTION_LIST_COUNT=2

CONFIG_AWS_IOT_TOPIC_UPDATE_DELTA_SUBSCRIBE=y

CONFIG_AWS_IOT_LAST_WILL=y

CONFIG_AWS_IOT_TOPIC_GET_ACCEPTED_SUBSCRIBE=y

CONFIG_AWS_IOT_TOPIC_GET_REJECTED_SUBSCRIBE=y

RE: nrf9160 reboots sporadically after some time.

Didrik Rokhaug — Wed, 22 Mar 2023 13:55:50 GMT

Could you add CONFIG_TFM_SECURE_UART_SHARE_INSTANCE=y to your prj.conf as well?

That should let the secure fault handler output some information about the crash.

[quote user="J0sh1101"]What could be the reason to execute the secure fault handler?[/quote]

The application trying to access something in secure memory is the most likely cause. The reason for this might be a stack or buffer overflow, or another form of pointer error.

RE: nrf9160 reboots sporadically after some time.

J0sh1101 — Tue, 21 Mar 2023 16:32:17 GMT

Hi Didrik,

I added the flags to prj.conf and added a new LTE link monitor result in the first post. Basically, there is now the prj.conf file and the lte link monitor 2 file. From the logs, it appears that the application just stops and waits after it sends the second time. After debugging the application, I found that it was stuck in a secure error handler. I tried tracing the call stack to see what the error could be, but I couldn't dig deeper because the "source is unknown". Below is a snippet of my ide and the call stack.

snipped of the IDE:

What could be the reason to execute the secure fault handler?

RE: nrf9160 reboots sporadically after some time.

Didrik Rokhaug — Tue, 21 Mar 2023 12:41:42 GMT

Hi,

The reboot is probably because the application crashes.

Could you add CONFIG_RESET_ON_FATAL_ERROR=n and CONFIG_THREAD_NAME=y to your prj.conf?

That should give us more information about why the application crashes.

Best regards,

Didrik