Accredited Standards Committee*
InterNational Committee for Information Technology Standards (INCITS)
Doc. No.: T10/05-016r0
Date: December 17, 2004
Reply to: John Lohmeyer
To: T10 Membership
From: Mark Evans and John Lohmeyer
Subject: SAT Working Group -- December 16th and 17th, 2004
Las Vegas, NV
Agenda
1. Opening Remarks
2. Approval of Agenda
3. Attendance and Membership
4. Old Business
4.1 SAT INQUIRY contents (04-218r7) [Elliott]
4.2 ATA Command Pass-Through (04-262r2) [Stevens]
4.3 SAT: Remaining Work Items (04-391r1) [Sheffield]
5. New Business
5.1 SAT ATA Protocol Specific Mode Page Proposal (04-397r0) [Overby]
6. Review of SAT Working Draft
7. Review of Recommendations
8. Meeting Schedule
9. Adjournment
Results of Meeting
1. Opening Remarks
John Lohmeyer called the meeting to order at 1:00 PM Thursday, December 16,
2004. He thanked Dan Colegrove of Hitachi GST for hosting the meeting. As
usual, the people present introduced themselves.
2. Approval of Agenda
The draft agenda was approved with no changes.
No items were added or revised during the course of the meeting.
3. Attendance and Membership
Attendance at working group meetings does not count toward minimum attendance
requirements for T10 membership. Working group meetings are open to any person
or organization directly and materially affected by T10's scope of work. The
following people attended the meeting:
Name S Organization
---------------------- -- -------------------------
Mr. Ralph O. Weber P ENDL Texas
Mr. Wayne Bellamy V Hewlett Packard Co.
Mr. Dan Colegrove P Hitachi Global Storage
Mr. Steve Livaccari V IBM Corp.
Mr. Nathan Marushak V Intel Corporation
Mr. Robert Sheffield P Intel Corp.
Mr. John Lohmeyer P LSI Logic Corp.
Mr. Steve Johnson V LSI Logic Corp.
Mr. Mark Evans P Maxtor Corp.
Mr. Mark Overby V nVidia Corp.
Mr. Jim Hatfield V Seagate Technology
Mr. William Martin V Sierra Logic
Mr. Curtis Stevens P Western Digital
13 People Present
Status Key: P - Principal
A,A# - Alternate
AV - Advisory Member
L - Liaison
V - Visitor
4. Old Business
4.1 SAT INQUIRY contents (04-218r7) [Elliott]
In Rob Elliott's absence, Wayne Bellamy reviewed revision 7 of this proposal
about data that is returned for an INQUIRY command.
There was much discussion about names and topology, for example, target device
names derived from model number and serial number. The proposed identifier
type of 1 (T10 vendor identification) for target device names would allow
non-globally unique names. One possibility is: if a WWN is provided by a
SATA device in its IDENTIFY DEVICE information, that WWN could be used as the
target port identifier. If a WWN is not provided by a SATA device, the target
port identifier could be the expander port name.
During the discussion Ralph Weber stated that George Penokie intended that
only target devices containing well-known logical units be required to have
target device names. Wayne said that he would remove the clauses in his
proposal concerning target device name.
Wayne moved that 04-218r8 (r7 as revised based on input from the working
group) be recommended for inclusion into the SAT draft. Mark Evans seconded
the motion. The motion carried unanimously (8:0:0).
4.2 ATA Command Pass-Through (04-262r2) [Stevens]
Curtis Stevens reported that he received the op codes from the CAP Working
Group to define new SCSI CDBs for issuing ATA commands and sense information
to report completion status. He will incorporate these into the next revision
of the proposal.
4.3 SAT: Remaining Work Items (04-391r1) [Sheffield]
Bob reviewed revision 1 of the list that he is maintaining that itemizes the
known work items to be completed for the project and the status of those
items. Bob reported that he will continue to maintain this list until the
project is complete.
5. New Business
5.1 SAT ATA Protocol Specific Mode Page Proposal (04-397r0) [Overby]
Mark Overby reviewed his proposal for a protocol specific mode page for SAT.
There are two subpages, one for parallel ATA devices and one for serial ATA
devices.
It was recommended that Mark increase the size of the pages to provide more
reserved bytes. Mark agreed to make this change.
It was suggested that these might be better as Control Mode subpages. Mark
agreed to make this change.
There was a comment that information presented in the SATA subpage (e.g.,
status) was not available. Mark said that he would check to see why he
included this.
Mark agreed to make a new revision of his proposal including input from the
working group.
6. Review of SAT Working Draft
Bob Sheffield began by reviewing change bars and editor's notes in the draft.
Mark Overby stated that the definition for "ATA domain" ("An I/O system
consisting of a set of ATA hosts and ATA devices that communicate with one
another by means of a service delivery subsystem.") is incorrect. Mark stated
that there can only be one host in an ATA domain. Ralph Weber stated that the
current definition in the draft is consistent with the definition in the SAS
standard and that Mark's definition of an ATA domain (i.e., having only one
host) conflicted with the concept of a SCSI domain. Mark Evans stated that
Mark Overby's definition is correct and is consistent with the definition in
the preliminary version of the ATA Architecture Model draft standard.
Mark Evans agreed to review the SAS draft standard and propose wording to
change the definition of "ATA domain" (and possibly other related definitions)
to be correct and consistent.
There was more discussion about power management (e.g., START STOP UNIT).
Mark Evans agreed to review the draft for consistency between SPC and
ATA/ATAPI standards.
There was extended discussion about the REASSIGN BLOCKS command mapping.
Several folks stated that there was no way to reassign a block on an ATA
device after a write failure because an ATA drive would only report a write
error (via the DF bit in the Error register) after there were no more spare
LBAs. It was also stated that an ATA drive would only consistently reassign
an LBA after a read failure when the LBA was written. Therefore, it could be
possible to reassign an LBA by performing a READ VERIFY. If the READ VERIFY
succeeds then the LBA has been reassigned if necessary. If the READ VERIFY
fails, then issue a WRITE command to that LBA and the device will reassign the
block if necessary. The WRITE command should be followed by a READ VERIFY to
insure that the LBA was written correctly.
Dan Colegrove agreed to create an agenda item for the next T13 meeting to
review this reassignment algorithm with that group.
The next discussion was about mapping of the FORMAT UNIT command. It was
stated that ATA devices could not support any form of defect list. It was
noted that SBC-2 only requires a vendor-specific defect list format, which
could be no defect list. Also, SBC-2 does not require the medium to be
altered by a FORMAT UNIT command. Ralph Weber recommended, and all agreed,
that SAT require only the vendor-specific defect list format option with no
defect list. If an application client requires that the media be written,
then this function could be mapped to using the ATA WRITE SAME command.
Bob stated that he will not make a new revision of this document until we
complete review of the current revision of the draft.
7. Review of Recommendations
Mark Evans noted that the working group made the following recommendation to
the T10 plenary:
For inclusion in SAT: SAT INQUIRY contents (04-218r8) [Elliott], {r7
as modified, unanimous}
8. Meeting Schedule
SAT Working Group meetings are scheduled for:
Tuesday, January 18, 2005 from 9:00 AM to 12:00 PM in Las Vegas, NV at the
Tropicana Hotel (800-634-
4000), hosted by Hitachi GST.
Thursday, February 25, 2005 commencing at 1:00 PM ending at 12:00 PM on
Friday, February 25, 2005. The meeting will be held at the nVidia facility in
Santa Clara, CA. Mark Overby will send a list of recommended hotels to the
reflector.
Teleconference calls may be announced on the T10 reflector as needed.
9. Adjournment
The meeting was adjourned at 11:30 AM on Friday, December 17, 2004.