TMS570LC4357: ESM High Interrupt occurs with no pending interrupt in ESMIOFFHR

Hi Cameron,

Can you please also check all the ESM status registers also whether any of them got SET?

To verify this set the breakpoint at the beginning of the esmHighInterrupt function.

And also, please refer my below threads:

(+) TMS570LS3137: I am planning to perform a selftest using the SL-selftest function provided by the TMS570LS3137 chip. - Arm-based microcontrollers forum - Arm-based microcontrollers - TI E2E support forums

(+) TMS570LC4357: SDL ESM Configuration (Error Signaling Module) - Arm-based microcontrollers forum - Arm-based microcontrollers - TI E2E support forums

Here i mentioned how to verify the which ESM channel causing the interrupt by verifying the argument passed to the notification function.

--
Thanks & regards,
Jagadish.

Hi Jagadish,

All the ESM Status Registers read 0x00000000 at the breakpoint I set at the beginning of esmHighInterrupt.

Neither esmGroup1Notification nor esmGroup2Notification is called, because esmREG->IOFFHR equals 0, so the value of vec is set to 0xFFFFFFFF and the else path is taken.

Do you have any other ideas?

Thanks,

Cameron

Hi Cameron,

After coming into esmHighInterrupt, to which of the below if or else condition the code is entering?

--
Thanks & regards,
Jagadish.

It is hitting the last else:

else
{
esmREG->SR7[0U] = 0xFFFFFFFFU;
esmREG->SR4[0U] = 0xFFFFFFFFU;
esmREG->SR1[1U] = 0xFFFFFFFFU;
esmREG->SR1[0U] = 0xFFFFFFFFU;
}

Hi Cameron,

I don't know why we are getting esmHighInterrupt without any ESM status flags?

Is it possible to share your code for debugging? or else any simplest project? You can send through private chat as well.

--
Thanks & regards,
Jagadish.

Hi again Jagadish,

I am not able to share my code, but as an experiment I tried running it on three other devices, and the ESM High Interrupt did not occur on any of them. I am beginning to suspect there may be a faulty MCU on the device I was using, although it seems odd that a rogue interrupt on startup would be the only symptom.

Thanks,

Cameron 

Hi Cameron,

Cameron Fruit said:

I am not able to share my code, but as an experiment I tried running it on three other devices, and the ESM High Interrupt did not occur on any of them

That was strange.

Cameron Fruit said:

I am beginning to suspect there may be a faulty MCU on the device I was using, although it seems odd that a rogue interrupt on startup would be the only symptom.

Maybe you are correct.

If possible, send your code, so that i can test it once at my end. This will give some confidence about the code.

--
Thanks & regards,
Jagadish.

Hi Jagadish,

Sorry for the late reply as I was out of the office the last few days.

I can't send my code, but we are checking out the hardware to see if there is an issue there.

Thanks,

Cameron

Hi Cameron,

Cameron Fruit said:

we are checking out the hardware to see if there is an issue there.

Sure, check and let me know, mean time i will also try to verify other possibilities for this issue.

--
Thanks & regards,
Jagadish.

The hardware appears to check out fine, but I'm still seeing the ESM High Interrupt handler invoked with clear ESM status registers and clear ESM High Offset register. Are there any other conditions that can cause an ESM High Interrupt?

Thanks,

Cameron

Hi Cameron,

As per as i know, every ESM interrupt should set the corresponding bit in the status flags, i don't think any interrupt will generate without status bits.

Can we set one live debugging session, as this issue is not triggering at my end?

I will be available from 10AM to 7PM IST (Indian Standard Time), so based on your convenient please set one live debugging session for better understanding of issue and to provide my suggestions.

--
Thanks & regards,
Jagadish.

Hi Jagadish,

I am not able to have a live debugging session, but from what I can tell, the ESM High Interrupt occurs before the first statement in the c_int00() function. Does this suggest anything to you?

Thanks,

Cameron

Hi Cameron,

I never see this kind of issue before,

I will try to check with one of my senior colleagues whether he got any of this kind of issue before.

--
Thanks & regards,
Jagadish.