AVR-Chat

At 07:16 AM 2/12/2004 -0800, Larry Barello wrote:
>You are describing a classic race condition.  I think Brian Dean is right on
>the money where the problem is.  Race conditions are very hard to smoke out.
>You need a reliable trigger - or, if that isn't possible, then a repeatable
>test case (which is the same thing).  Since you have two CPUs the only way
>to get it repeatable is to synchronies the clocks & reset.

At this point, I'm going to head home, take a couple days off, and then 
attack it again fresh.

It is proving a challenge to find.