Re: Random IDE Lock ups with via IDE

[Ross Biro]

If you are using the legacy IDE layer you want to tweak WAIT_CMD.  For
testing, you can make it really large and see if that impacts your
problem.  A drive vendor once told me that it could take more than a
minute for an IDE drive to complete a command.  I no longer purchase
drives from that vendor.

I'm not sure what libata uses, but my guess is it defaults it from the
SCSI layer.

If tweaking these time outs make your problem go away, odds are what
happened was that your drive remapped a few more bad sectors and now
takes a little too long to complete commands.  The linux ide error
recovery code does a WIN_IDLE_IMMEDIATE when there is a problem.  This
is allowed by the ATA-2 spec, but confuses most modern drives.  So
once you start getting errors, often the drive gets so confused, you
never stop.

    Ross

On 5/26/05, Jim Gifford <maillist@xxxxxxxxx> wrote:
> What do you recommend trying?
>
-
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/