Recover ChipSHOUTER from Max_Retry_Exception

Embedded Security Glitching
1

We are using the ChipSHOUTER in our lab, but for long runs we experience the ChipSHOUTER raising the Max_Retry_Exception error.

In general, we recover by reseting the CS and rewriting all our parameters. In this case it doesn’t work and we have to restart our framework, which closes and reopens the connection to the CS.

Are there any better ways to recover the CS from such an error?

Thanks! Best regards,
Alex

2

Hi Alex,

What firmware version are you on? We did have some improvements in later versions (latest is 2.0.1 from memory) to reduce resets / lock-ups.

But it sounds like the ChipSHOUTER is locking up or resetting. We’ve found that does happen in various setting combinations, which is why we recommend recovering via the reset method when needed. More often we see the ChipSHOUTER itself reset.

When you get that exception is anything unusual happening, what are you sending it?

Generally we recommend keeping as much state “externally” as possible so you can recover as you suggest, so that might be the “best” way. The ChipSHOUTER was by design allowed to run in areas which can cause internal stability, as those “unstable” areas vary per unit. So rather than add artificial limitations we let you push the limits of it as far as possible (and beyond). The “safety limits” where damage occurs are nowhere near this, so basically there isn’t a risk of much beyond the annoyance of the lock-up/reset problem.

I’ll ping @Alex_Dewar if he’s seen this issue as well, maybe there is a better fix than I know of!

3

FW version is 1.8.7
So maybe a little bit out of date. Do I need to send a mail to supoort with the board id?
(If so, I will do this tomorrow as I don’t have remote access to all our CSs.)

When you get that exception is anything unusual happening, what are you sending it?

Not much, every 5 secods trigger_safe is checked and in between shout several times using an Hardware Trigger.

4

Yes - we’re still manually building sorry, so we need the board ID so it matches!

Lets see what happens with the latest FW version, there were a few fixes in there related to reliability.

5

If it’s easier, feel free to just DM me on the forums

Alex

6

Thanks for the FW updates. This has definitely changed behaviour for the better. We no longer experienced Max_Retry_Exception or other errors like IOError.

1 Like