Re: [BUG?] lost interrupt

[Måns Rullgård]

David Caldwell <david+kernel@xxxxxxxxxxxx> writes:

>> With all 2.6.0 versions so far, I get these errors when writing lots
>> of data to the disk:
>
> I am getting these same errors with 2.6.0-test6. The difference is,
> I'm not using a SiS IDE controller. I have a Promise 20276 on my
> motherboard which was the controller getting the lost interrupt
> error. I am running RAID5 using disks on this controller and on a
> Promise 20267 PCI card (note: 67 not 76!). I seemed to start getting
> the error when my disks started going with lots of activity.
>
> I had booted with "noapic" at the time. Without "noapic" my
> motherboard wouldn't boot at all (it seemed to hang right after
> detecting the IDE devices, but I don't know if that is relevant).

That doesn't make any difference for me.

>> Losing too many ticks!
>> TSC cannot be used as a timesource. (Are you running with SpeedStep?)
>> Falling back to a sane timesource.
>
> I was definitely getting this same error.
>
>> hda: lost interrupt
>> hda: lost interrupt
>> hda: lost interrupt
>> hda: lost interrupt
>
> I was definitely getting this error, except it was on hdg (my promise
> 20276).
>
>> hda: status error: status=0x58 { DriveReady SeekComplete DataRequest }
>> 
>> hda: drive not ready for command
>
> I don't remember if I was getting this error or not.
>
>
> I didn't have the forethought to save my dmesg output. This screwed up
> my RAID pretty bad, so I'm a little reticent about making it happen
> again...

In my case, no data is lost.  It just takes an awfully long time to
write it.

> I just wanted to let it known that it wasn't just happening to SiS IDE
> controllers.

Could you post your dmesg output for that machine?  It might reveal
something, even the error doesn't occur.

-- 
Måns Rullgård
mru@xxxxxxxxxxxx

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/