Humm, I don't suppose you can connect a scope to SDA and SCL to see what is happening? It sounds like a hardware flaw. :'( Phil Joe Drew wrote: >On Thu, 2002-12-19 at 15:40, Philip Edelbrock wrote: > > >>It's likely that a chip on the bus is hanging it. Is there a lead-up to >>this state? How can it be reproduced on your computer? >> >> > >Unfortunately, I know of no reliable way to reproduce it. This happens >fairly often, it seems, but it doesn't *always* happen, even when >diagnostics are running. The diagnostics simply run 'sensors' for a >particular chip every 5 minutes (and parse the output). > >If it makes a difference, this is on a Ziatech 'blade' CPU card - a >CompactPCI machine. It happens even on different model machines where >the only chip that is on the SMBus is a Maxim 1617A, and the only >relevant modules loaded are adm1021, i2c-piix4, and dmi_scan. > > >
