MPC5674F Censorship Password set from debugging

[ghgpozzato]

Hello NXP engineers, 

I was debugging an normal application in a MPC567XEVBMV with a MPC5674F core. Out of nowhere the debugger could not connect anymore and request for the censorship password. 

The application is running from flash but there are no attempts to change the shadow block.

I am using the S32DS Power PC IDE with an P&E Micro probe for debug. 

I already tried the default public password without success.

-uncensorpassword=FEEDFACECAFEBEEF in the GDB server parameters.

If you have any suggestion on what can be done to recover the MCU. In case it is locked forever, what can we do to prevent this scenario to happen again.

Kind regards 

[ghgpozzato]

Hello @lukaszadrapa,

Thanks for the fast response. 

I have a MPC5676R to replace the MPC5674F and when I do everything works perfect. I tried to clean and push the MCU into the socket without any success. I can also see the 3 LEDs in for 1.2V, 3.3V and 5V in the evaluation board are on so the MCU is at least being powered. Another thing is that the Reset LED (D11) is always on in the faulty MCU. On the working MCU it just blinks are power on.

Kind regards, 

Guilherme 

[lukaszadrapa]

Hmmm, hard to say what happened. But if MPC5676R works and MPC5674F hangs in reset forever, I guess that MPC5674F is damaged somehow. There are not many options here, so I would replace the device.

Boards with socket are great for testing, we can easily change the MCU as needed but those are not always 100% reliable from my experience.

Regards,

Lukas

[lukaszadrapa]

Hi @ghgpozzato 

this doesn't seem to be related to censorship. It looks rather like some HW issue. If censorship is enabled, the debugger will not get any response from the MCU and password needs to be provided. So, if something goes wrong, the debugger tries to uncensor the device.

If you have version of board with socket, there may be bad contact. Push the socket down, "shake" the MCU a little bit in the socket and release the socket. This usually helps.

Another problem is that ribbon cable between debug probe and board is bad.

Regards,

Lukas