TMS570LC4357: FEE driver reinitialization fails with watchDog failures on TMS570LC43

Hi Indula,

Which board you are using? TI-Specific or Custom?

I am suggesting you refer below FAQ once:

(+) [FAQ] TMS570LC4357: How to use Digital Watchdog - Arm-based microcontrollers forum - Arm-based microcontrollers - TI E2E support forums

In above thread i explained couple of problems regarding watchdog reset, please have a look on them once.

If it doesn't help then, could you please share your code for quick debugging of issue at my end?

--
Thanks & regards,
Jagadish.

HI Jagadish,

Sorry for the late reply, We use custom hardware and unfortunately I can`t share the code with you
But I will try to explain what we are trying to achieve.

We are using the Digital watchDog to trigger a system reset. SO hopefully that`s okay still even if the NMI has issues (we are not using NMI)

We also use the emulated EEPROM driver to store some information into flash Bank7 at a regular interval

If the system had an unexpected watchDog failure, we try to restore the system state into a last known state from the EEPROM

But sometimes, EEPROM driver returns busy after a watchDog failure and curious why this is the case

Thank you

Indula

Hi Indula,

In our old thread we found that initializations will not happen properly in watchdog reset condition. Please refer below thread for more details:

(19) TMS570LC4357: WDT handling - upon watchdog reset - Arm-based microcontrollers forum - Arm-based microcontrollers - TI E2E support forums

May be this is creating difference here.

My suggestion would be:

1. If you are using TI-Launchpad then do the below code modification on your board and do the test:

2. If you are using TI-HDK board then along with above medication you should also need to do below modification your code:

Just try these modifications and verify the behavior, please refer above shared thread for more details about these medications.

--
Thanks & regards,
Jagadish.

Hi Jagadish,

We are using a 3rd party RTOS and it does the system initialization upon boot. So I am not sure how I could modify the startup.

On a similar note, I have found that if you call TI_Fee_Init() more than once during boot it could sometimes lead to a data abort.

In our implementation, FEE globals stay in an external ram and during a watchDog reset that memory remains intact.
Maybe the watchDog causes the TI_Fee_Init() to be called twice as there is no power interruption happening with a watchDog reset ?

Thank you

Indula

Hi Jagadish,

I did more testing and confirmed that it was TI_Fee_Init() that get stuck (sometimes) when called more than once during initialization.
So I made sure to only call the function once and seems to be working as acceptable.

Thank you

Indula

HI Indula,

Indula Abeyrathne said:

I did more testing and confirmed that it was TI_Fee_Init() that get stuck (sometimes) when called more than once during initialization.
So I made sure to only call the function once and seems to be working as acceptable.

Good to hear that.

Good work!

--
Thanks & Regards,
Jagadish.