On Wed, Jun 16, 2004 at 11:26:02AM -0400, Reuben D. Budiardja wrote:
> On Wednesday 16 June 2004 11:01 am, Arjan van de Ven wrote:
> > On Wed, 2004-06-16 at 16:53, Reuben D. Budiardja wrote:
> > > On Wednesday 16 June 2004 10:14 am, you wrote:
> > > > > hdi: dma_intr: error=0x84 { DriveStatusError BadCRC }
> > > > BadCRC errors are cable faults 99.9% of the time.
>
> > > If it's broken, shouldn't I be having trouble all the time, not just
> > > during booting ?
> >
> > well the kernel will slow down the bus speed after such errors, so it
> > could well be that the initial bus speed is estimated too high, and
> > after the errors it scales down.
> > (but note: there comes a speed, eg sub udma, where there no longer is
> > CRC's with the data but just silent corruption instead)
>
> So should I worry about it, about corruption, or not ?
It's not a good sign.
do check with hdparm -i which dma mode is active... if it's udma then your
data at least is CRC protected.
> I guess I just don't understand why this happen only to one drive, out of 3
> drives with the same controller and cable. If the kernel estimated the wrong
well there's a maximum of 2 disks per cable...
Attachment:
pgp00009.pgp
Description: PGP signature