Data field size of intermediate data groups

Robert Frey bobf at advansys.com
Thu Mar 4 10:06:22 PST 1999


* From the T10 Reflector (t10 at symbios.com), posted by:
* bobf at advansys.com (Robert Frey)
*
Bill Galloway wrote:
> I do not look forward to supporting this either but I think we have to.  Any
> device that does not have a multiple of four block size could get into trouble.
> I could imagine a tape drive that wants/needs to do a CRC every block. If the
> block size was not a multiple of four then they would violate your rule.  Also
> not all disk are formatted for 512 bytes, I know of some that are formatted for
> 514. This disk could get into trouble with intermediate CRCs as well.

To extend this line of thought because there is no way to indicate a
1 or 3 byte pad field in a data group, then don't DT Data Phases
implicitly preclude odd byte record sizes. This maybe an unintended
consequence of assuming that CRC follows the (upper layer concept)
of a record size.

The transport layer and data bus protection/CRC would ideally
(I think) work independently of an upper layer record size. For
ST Data Phases the upper layer indictes a record size adjustment
with the IGNORE WIDE RESIDUE message, but there is no way to indicate
such an adjustment between DT Data Phase data groups.

Are target implementers in fact planning on linking the transport
CRC to the target record size?

Bob Frey
bobf at advansys.com
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at symbios.com





More information about the T10 mailing list