Device Identifiers in SMC Read Element Status

Erik Falk ErikF at
Thu Feb 26 18:09:42 PST 1998

* From the T10 (formerly SCSI) Reflector (t10 at, posted by:
* Erik Falk <ErikF at>
In SMC revision 10, the paragraph describing the IDENTIFIER field in the
definitions of every element descriptor (sections 6.5.3, 6.5.4, 6.5.5,
and 6.5.6) contains the following sentence:
"If no device identifier is available for this element, or the DVCID bit
in the CDB is zero, this field shall be omitted."

In addition, the first paragraph following table 16 in section 6.5.7
contains the following two sentences:
"Device identifiers may be available for some or all elements in a
medium changer. If no device identifier is available or the DVCID bit in
the CDB is zero, the IDENTIFIER LENGTH shall be zero, the IDENTIFIER
field is omitted, and the CODE SET and IDENTIFIER TYPE fields shall be

This seems to imply that not all the element descriptors for an element
type must be the same size (e.g. if one tape drive in an autoloader
supports device identifiers but a second doesn't, or their device
identifiers are of different sizes, the descriptors will have different
sizes).  If this is the case, what values should be used for the BYTE
COUNT OF REPORT AVAILABLE field of the element status data header and
ELEMENT DESCRIPTOR LENGTH fields of the element status page header?
Also, this would apparently require an application to parse every field
in every preceding element descriptor of the same type in order to find
the descriptor for a specific element address.

As a possible solution, the end of the vendor-specific area of each
element descriptor can be "padded" with zeroes so that all element
descriptors of the same type have the same size.

Please let me know if I am misinterpreting SMC in this regard, or else
if there is a "proper" solution to this problem.

Erik Falk
10201 Willows Road
Redmond, WA 98073
425-881-8004 x3444
erikf at

* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at

More information about the T10 mailing list