Re: corrupted data with ide-scsi emulation

Andre Hedrick (andre@suse.com)
Tue, 4 Jan 2000 00:04:42 -0800 (PST)


On Tue, 4 Jan 2000, Tuukka Toivonen wrote:

> I *always* get the right result. So could this be a kernel problem instead
> of hardware problem?

Yes-No.........

> Also, if I do
> dd if=/dev/cdrom bs=1024 count=xxx | cmp - original.image
> it succeeds also always.
>
> Also strange is that some (maybe 20%) of CD-R disks work fine and give
> correct md5sum even if running md5sum in pipe.
>
> I have had never problems while using the CD-R disks normally mounted,
> except when I tried to set unmaskirq to on with hdparm, which gave lot of
> read errors.

Well if you give way your interrupts (figuratively speaking) I bet you get
a buffer underrun......

> This is Mitsumi CR2801TE cd-r drive, and I have tried really many of 2.0.x

And the core engin for a "CR2801TE" is ??

> kernels as well as 2.2.13 which all give the same result.
> The cd-r drive is in /dev/hdc, my hard disks are in /dev/hda and hdb and
> /dev/hdd is not attached to anything.

Since you never defined a chipset base or cpu class, you have mad any
guess impossible.

Cheers,

Andre Hedrick
The Linux IDE guy

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