On Sun, Apr 13, 2008 at 9:23 PM, dean gaudet <dean@xxxxxxxxxx> wrote: > oh it finished, sorry i should have said that :) > > there were just those various soft lockup warnings... > > as a wild stab in the dark -- does the "faulty" test mode support > delaying reads or writes for a very long time? (or is there some other > fake block device we can inject long delays with?) > DM has an I/O delaying target... but that should not be the problem in this case. The 'softlockup' mechanism triggers when we are spending an inordinate amount of time *running* in a given kernel routine, while MD is waiting for I/O to come back the md thread is sleeping. I looked at the code and found a live lock condition, but if you had encountered it your array would be degraded and the resync process would have never completed. The fact that it is triggering in async_xor_zero_sum makes me wonder if this could be a false positive i.e. a perfect storm where the softlockup tick is always happening to fire while we are computing parity. Seems improbable though... I would need to understand more about how softlockup works before making that assertion. Thanks, Dan -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
