Diagnosis for an unusual TRAP occuring only on power-up?

[Ridgeglider]

We have been experiencing unusual Trap errors on a 5234 connected to quite a bit of external hardware. I say unusual because the code runs fine on the basic development boards (w/o external hardware). It also runs fine on the 5234 connected to the external hardware: We can download and run the problematic code image without problems, and we can download new images when the culprit code is loaded. The only time a trap occurs is when we switch off, and then power-up. We're obviously lookiong as the (non-NB) power supply during shutdown and powerup, but any other suggestions?

Our external hardware does not include memory-mapped devices using any of the CS lines. It does include the SDcard, about 5 or 6 serial ports, and quite a few of the 5234 IO pins, most as gpio, but some frequency counters.

Thanks

[tod]

I have no idea what you've tried so please forgive me if what I suggest is too simple.

1. Have you tried lengthing the startup delay for the NB?
2. Do you have smart traps enabled? Is the trap occuring in a repeatable location?
3. Since you are looking at the PS make sure you look at the voltage going to the NB, we once had traps that occured because we just drew power over the limit causing the voltage to drop below the threshold to the NB and it would restart.

Tod

[Ridgeglider]

Hi Tod: Thanks for your response. We figured it out. One of us had an earlier NB version. Most of us had Rel22rc2, but one of us still had Rel22rc1. The way we noticed was that Rel22rc2 prints out the build info to the console when it's done, while the earlier rev does not. (Thanks Paul!). This could have taken a LOT longer to figure out without that difference.

Interestingly, the rel22rc1 still built and ran the rel22rc2 code when downloaded as long as we didn't power down and back up. I'm sure it would have gotten us somewhere else though, but just something to be aware of.

Thanks for offering help.