Tejun wrote: Yeap, I just did. I'm looking at ATAPI_7_LiteOn.sas. The offending one is the one ending with the FIS 22.723 807 202, right? On the bus level, everything seems fine. The host prepares for 96bytes transfer and device does it and successfully sends D2H Reg FIS for command completion. The only thing I can think of is the draining buffer confusing the controller somehow. If you make libata-scsi.c::atapi_drain_needed() always return zero, does anything change? ++++++++++++++++++++ Tejun, The problem occurs in 2.6.18.8 which doesn't have atapi_drain_needed() and in 2.6.28.7 which does. I had atapi_drain_needed() return zero on 2.6.28.7 with no change in behavior. If you click the ATA button (ATA application layer) on the above trace and expand the IN and OUT boxes on transaction 22.723 597 590, you see the DRDY, SERV, I/O, C/D pattern that I described before. Does this pattern make sense in this context? Do you think that this pattern is of concern or should we be looking elsewhere for the problem? Doug -- To unsubscribe from this list: send the line "unsubscribe linux-ide" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
