additional sense codes for overlapped tags

Gerry.Houlder at seagate.com Gerry.Houlder at seagate.com
Mon Jan 7 15:53:27 PST 2002


* From the T10 Reflector (t10 at t10.org), posted by:
* Gerry.Houlder at seagate.com
*
* From the T10 Reflector (t10 at t10.org), posted by:
* Gerry.Houlder at seagate.com
*

My impression is that checking for overlapped commands (e.g. duplicate tag
values) was optional. FC products in particular didn't want to add this
overhead. Has it suddenly become required again?



                                                                                               
                    "Elliott, Robert"                                                          
                    <Robert.Elliott at C        To:     <t10 at t10.org>                             
                    OMPAQ.com>               cc:                                               
                    Sent by:                 Subject:     additional sense codes for           
                    owner-t10 at t10.org        overlapped tags                                   
                                                                                               
                                                                                               
                    01/07/02 04:51 PM                                                          
                                                                                               
                                                                                               




* From the T10 Reflector (t10 at t10.org), posted by:
* "Elliott, Robert" <Robert.Elliott at compaq.com>
*
A minor issue was raised in today's SRP call:
Does a multi-byte tag field need to include MSB and LSB labels?

Ed searched through SAM-2 and SPC-3 and found only one place where
the tag is interpreted as a numeric value, making this distinction
necessary.  When the target detects an overlapped tag it needs to
choose which additional sense code to return:
  4Dh/nnh if the tag is FFh or less;
  4Eh/00h if the tag is greater than FFh.

SCSI-2 targets always returned 4Eh/00h, so this is one of the
elusive differences between SCSI-2 and SCSI-3.

SCSI-3 excerpts
===============
SAM-2 5.8.2 Overlapped commands
  A task manager that detects an overlapped command shall abort all
  tasks for the faulted initiator in the task set and the device
  server shall return CHECK CONDITION status for that command. If
  the overlapped command condition was caused by an untagged task
  or a tagged task with a tag value exceeding FFh, then the sense
  key shall be set to ABORTED COMMAND and the additional sense code
  shall be set to OVERLAPPED COMMANDS ATTEMPTED. Otherwise, an
  additional sense code of TAGGED OVERLAPPED TASKS shall be
  returned with the ADDITIONAL SENSE CODE QUALIFIER field set to
  the value of the duplicate tag.

SPC-3 additional sense codes table:
  4Dh NNh TAGGED OVERLAPPED COMMANDS (NN = QUEUE TAG)
  4Eh 00h OVERLAPPED COMMANDS ATTEMPTED

SCSI-2 excerpts
===============
SCSI-2 7.5.2 Incorrect initiator connection:
  A target that detects an incorrect initiator connection shall abort
  all I/O processes for the initiator on the logical unit or target
  routine and shall return CHECK CONDITION status. The sense key shall
  be set to ABORTED COMMAND and the additional sense code shall be
  set to OVERLAPPED COMMANDS ATTEMPTED.

SCSI-2 additional sense code table:
  4D 00
  4E 00 OVERLAPPED COMMANDS ATTEMPTED
  ALL CODES NOT SHOWN OR BLANK ARE RESERVED.

SCSI-3 comments
===============
The SCSI-3 behavior may be useful for non-packetized parallel SCSI
where the REQUEST SENSE command is issued to retrieve the sense
data, and the tag is limited by the protocol to 8 bits.

It is not useful for autosense protocols, however, because the
response IU carrying the sense data already includes the full
tag.  Despite that, targets using these protocols are required
to return 4Dh/nnh rather than 4Eh/00h if the tag falls in a
certain numeric range (00h to FFh).

SCSI-3 and FCP-2
================
In FCP-2, there doesn't seem to be a proper definition of tag
that would clarify when the tag "exceeds FFh":

3.1.55 tag: ... The fully qualified exchange identifier performs
the function of the SCSI tag in this standard.
3.1.23 fully qualified exchange identifier: A set of addresses
and values used to uniquely identify an FCP I/O operation. See 5.1.
5.1 ... The FQXID is composed of the initiator address identifier,
the target address identifier, the OX_ID and the RX_ID. Other
definitions of FQXID are outside the scope of this standard. The
method used to identify FCP I/O operations internal to the
application client and the device server is not defined by
this standard.

Where are the MSB and LSB in a FQXID?  Do any FCP-2 devices
return 4Dh/nnh or do they all return 4Eh/00h?

Suggestions
===========
* Get rid of 4Dh/nnh altogether
* Make 4Dh/nnh only usable on SPI-4 with information units disabled
* Make 4Dh/nnh only usable on "protocols that have tag fields
less than or equal to 8 bits"

Any opinions?

---
Rob Elliott, Compaq Server Storage
Robert.Elliott at compaq.com


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



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




More information about the T10 mailing list