Mia Magnusson wrote: > If it can store at least a bit of data before the trigger, how about > just trigger it on the bank outputs of the CPU being anything else than > bank 15? That would be too early. We don't want to find what heppens when the CPU _starts_ testing the RAM. We want to find out what caused the CPU to _stop_ testing the RAM - that's when the error occurs. And we don't know when it really occurs. For what it's worth, it could be testing 65535 bytes correctly and bail out on the 65536-th (not that I consider it likely, but you don't really know). Regards, Michau.Received on 2018-07-04 11:00:05
Archive generated by hypermail 2.2.0.