Justin> Your problem is actually quite different. In your case, the timeout Justin> occurs in a data-out phase and the chips status tells us:

Justin> 1) The target still has us in data phase.

Justin> 2) SCSI dma is still enabled.

Justin> 3) Our data fifo is full of data to send on the SCSI bus, Justin> but its not going anywhere.

Justin> So, the aic7890 believes that the target has stopped requesting Justin> more data, but we still haven't changed phase. This is the classic Justin> symptom of a flaky bus where the signals for clocking data are Justin> degraded. The target probably didn't see one or more of the
controller's Justin> ACKs, so we are stuck. Double check your termination, cabling, power Justin> supply, etc, etc.

Thanks for the comments. I just checked once again, and everything seems to be in order. I decided to ax the active terminator and run in S/E mode which has been working fine for >24h. Crossing my fingers ...

To Unsubscribe: send mail to majo...@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message