AM2434: AM2434 RTI-Watchdog reset reason

Ahmed Ali

Part Number: AM2434

I need to identify if the watchdog caused the reset.

the case that when the watchdog expires, it reports the error to the ESM and the ESM caused the reset, so the reset reason we are able to identify is ESM error.

I need to identify which esm error caused the reset, is it RTI-8, RTI-9,RTI-10, or RTI-11, or any other fault?

7 months ago

0 Johnson John 7 months ago

TI__Intellectual 2760 points

Hi,

The assigned expert is currently out of office. Please expect a reply in 1-2 days.

Thanks for your patience

Regards,
Johnson

0 Swargam Anil 7 months ago in reply to Johnson John

TI__Guru 54458 points

Hello Ahmed Ali,

Ahmed Ali said:
I need to identify which esm error caused the reset, is it RTI-8, RTI-9,RTI-10, or RTI-11, or any other fault?

Unfortunately, this is not supported on the SOC.

You can only monitor the ESM error, and this could be related to different RTIs .

Regards,

Anil.

0 Ahmed Ali 7 months ago in reply to Swargam Anil

Prodigy 10 points

But can i monitor the interrupt source and save it before reset or not? i mean in the esm callback or any solution

0 Swargam Anil 7 months ago in reply to Ahmed Ali

TI__Guru 54458 points

Hello Ahmed Ali,

Actually, the ESM interrupt is also a generic interrupt source.

When a core hangs in an exception state, the RTI is waiting for its timeout to occur. Once the timeout expires, the RTI attempts to trigger an ESM interrupt. However, since the core is already stuck handling the high-priority exception, the ESM interrupt cannot be serviced.

In this scenario, the RTI interrupt will never get executed. Therefore, instead of relying on logging the RTI interrupt, we need to use the SoC's AUTO-RESET feature to ensure system recovery.

Logging which R5F core initiated the reset is a little tricky. You can use IPC between all cores.

One core should monitor all IPC messages, and if any core hangs, the IPC notification message will not be received.

In this case, there might be a chance that the SoC gets reset when the corresponding RTI timer expires. However, if two cores hang at the same position, it is difficult to determine which core actually triggered the reset.

The above is my suggestion; if you really want to monitor the status of which RTI may trigger the SoC reset, then go ahead with the above proposal.

Other than this, we cannot use any other methods.

Regards,

Anil.

0 Ahmed Ali 7 months ago in reply to Swargam Anil

Prodigy 10 points

But how to use Soc's AUTO-RESET feature, i tried to configure the RTI reaction to RESET but that doesn't work, and it is mentioned that it is not supported

and if it is possible is there any example?

0 Swargam Anil 7 months ago in reply to Ahmed Ali

TI__Guru 54458 points

Hello Ahmed Ali,

Please look at the FAQ below .

[FAQ] AM64X/ AM62X : How to Reset the SOC when WDT timer expires in AM64X and AM62X?

Regards,

Anil.

0 Ahmed Ali 7 months ago in reply to Swargam Anil

Prodigy 10 points

if i configure the RTI in interrupt mode and assign it as the highest priority interrupt, will it possible that this interrupt will not be handled under any condition?

0 Swargam Anil 7 months ago in reply to Ahmed Ali

TI__Guru 54458 points

Hello Ahmed Ali,

If the core hangs in the exception and the RTI is configured to the highest priority, then even if the RTI timer expires, you don't get an RTI interrupt because the exception always has the highest priority.

Regards,

Anil.

Arm-based microcontrollers

Arm-based microcontrollers forum

AM2434: AM2434 RTI-Watchdog reset reason