[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: ide fails



Hi,

> hdc is less then a year old, had an hitachi (and replacement from
> hitachi) that died in that place twice. Both drives have a
> temperature about 38-42°C over the year. Yesterday I found that a
> package was brocken an elf binary wasn't in the right place. As a new
> kernel was on the update list, I rebooted the computer, fsck kicked
> in and was done with -fc options ... which found some bad blocks on
> hda. Fixed. Packages fixed, too. Those I noticed.
>
> And a day later I found on the console:
>
> Oct 21 07:20:55 platon kernel: hda: dma_intr: status=0xd0 { Busy }
> Oct 21 07:20:55 platon kernel: ide: failed opcode was: unknown
> Oct 21 07:20:55 platon kernel: hda: DMA disabled
> Oct 21 07:20:55 platon kernel: ide0: reset: success
> ...
> Oct 21 11:20:52 platon kernel: hdc: dma_intr: status=0x58
> { DriveReady SeekComplete DataRequest }
> Oct 21 11:20:52 platon kernel: ide: failed opcode was: unknown
> Oct 21 11:20:52 platon kernel: hdc: dma_intr: status=0x58
> { DriveReady SeekComplete DataRequest }
> Oct 21 11:20:52 platon kernel: ide: failed opcode was: unknown
> Oct 21 11:20:52 platon kernel: hdc: dma_intr: status=0x58
> { DriveReady SeekComplete DataRequest }
> Oct 21 11:20:52 platon kernel: ide: failed opcode was: unknown
> Oct 21 11:20:52 platon kernel: hdc: dma_intr: status=0xd0 { Busy }
> Oct 21 11:20:52 platon kernel: ide: failed opcode was: unknown
> Oct 21 11:20:52 platon kernel: hdc: DMA disabled
> Oct 21 11:20:52 platon kernel: ide1: reset: success
>
> I know that drives break. It's possible that two drives break at the
> same time. But 4 drives in all and two at the same time? Is there a
> problem with the PMac G3 ide? Do I need a custom kernel? Sarge was up
> over a year without a problem.

Try smartctl to see what the drivers have to report on their sanity. If
that does not show anything problematic, I'd suspect a driver bug or the
IDE interface is on its way out. If you have only ever been using the same
kernel, the driver bug excuse is out ...

The drives on a PMac are not byte swapped or anything out of the ordinary,
so transfering them to an external FW or USB drive bay could also help to
decide whether the drives or the interface are to blame.

	Michael




Reply to: