Accredited Standards Committee*
InterNational Committee for Information Technology Standards (INCITS)
Doc. No.: T13/05-155r0
Date: July 06, 2005
Reply to: Mark Overby
To: T13 Membership
From: Mark Overby
Subject: T13 Plenary Meeting (Irvine, CA) -- June 21-23, 2005
Agenda
1. Opening Remarks
2. Approval of Agenda
3. Attendance and Membership
3.1 Membership
3.2 Fees
3.3 Attendance
4. Approval of the Previous Minutes
5. Document Distribution
6. Review of Previous Action Items
7. Old Business
7.1 ATA/ATAPI-7
7.2 EDD-3
7.3 SCT Technical Report
7.4 HBA Standard
7.5 ATA/ATAPI-8 Projects
7.6 DRQ=1 when ERR=0 Technical Report Project
7.7 A/V Technical Report Project
7.8 T13 Reflector and Website Setup
8. New Business
8.1 INCITS Membership and Reflector Policy
8.2 International Participation by T13
8.3 HBA-2 Project Proposal
8.4 Items from the Reflector
9. International Representative Reports
10. Liaison Reports
10.1 T10
10.2 T11
10.3 Trusted Computing Group
10.4 SFF Committee
10.5 IDEMA
10.6 MMCA
10.7 CFA
10.8 SATA-IO
11. Call for Patents
12. Review of Action Items
13. Meeting Schedule
13.1 Meeting Length
13.2 Authorization of Ad-Hoc Working Groups
13.3 Long Term Meeting Schedule
14. Adjournment
Results of Meeting
1 Opening Remarks
Dan Colegrove, the T13 Chair, called the meeting to order at 0900 Tuesday, June 21st, 2005. Dan Colegrove thanked Curtis Stevens of Western Digital who hosted this meeting.
As is customary, the people in attendance introduced themselves, and a copy of the attendance list was circulated. Copies of the agenda were made available to those in attendance.
The Chair briefly reviewed the meeting rules: T13 is a Technical Committee of INCITS and operates under INCITS rules. Anyone with interest in the work of the committee is free to attend and take part in discussions. However, motions and votes are limited to member organizations, one vote per member organization.
The Chair also noted that T13 operates under the INCITS antitrust guidelines. There should never be discussion of the following topics at any INCITS or an INCITS subgroup meeting:
“Any company's prices or pricing policies;
“Specific R&D, sales and marketing plans;
“Any company's confidential product, product development or production strategies;
“Whether certain suppliers or customers will be served;
“Prices paid to input sources; or
“Complaints about individual firms or other actions that might tend to hinder a competitor in any market.
Policies and Procedures and antitrust guidelines for INCITS are available at the web site at www.incits.org. T13 operates under these policies and procedures augmented by the T13 procedures at the T13 web site at www.t13.org.
2 Approval of Agenda
Numerous corrections or modifications were made to the agenda and the agenda was republished with the corrections and additions. Mark Overby moved, Jim Hatfield seconded, that the agenda be approved as revised. The motion passed by unanimous consent.
3 Attendance and Membership
The Chair reviewed the rules for membership and attendance:
3.1 Membership
An individual from an organization new to T13 must attend at least one T13 plenary meeting before requesting voting membership. After that meeting, the representative may request membership from the T13 chair stating expertise and material interest in the scope of the committee and pay the required fees to INCITS. The time when membership becomes effective has just changed. Membership now becomes effective at the start of the next meeting the organization attends assuming the organization attends one of the next two meetings.
3.2 Fees
Current INCITS memberships fees for Principal or additional Alternates are 800.00 USD. The First Alternate is free with a Principal membership.
Mailings of committee documents are available to any interested party after each plenary meeting. The cost for mailings for one year is 300.00 USD for mailing in the USA and 400.00 USD for mailing to a location outside of the USA.
Persons with questions or requests regarding membership fees and/or subscriptions for mailings should contact Parthenia Purnell, ITI Standards Operations, at .
3.3 Attendance
In addition to payment of fees, there are also minimum attendance and participation requirements to maintain voting membership status on T13. A representative from a member organization must attend two of every three plenary meetings and vote on at least every other letter ballot. If a member organization fails to have a representative in attendance at two of the last three T13 plenary meetings or vote on every other letter ballot, they are warned that failing to attend the next T13 meeting or to vote on the next letter ballot will result in their voting membership being terminated.
Attendance at a working group meeting does not affect voting membership status in T13.
The T13 voting membership during the meeting was 17 organizations.
The number of eligible voting members required for establishing a quorum was 7. 15 Member organizations were present at the start of the meeting, thus a quorum was present.
The following individuals attended:
Ron Roberts Broadcom
Kenneth Hirata Emulex
Eric Lanning Finisar
Michael Lawson Finisar
Sumit Puri Fujitsu
Wayne Bellamy HP
Knut Grimsrud Intel
Robert Strong Intel
Brian Dees Intel
Robert Sheffield Intel
Tim Thompson LSI
Avraham Shimor M-System Flash Disk
Mark Evans Maxtor
Frank Shu Microsoft
Nathan Obr Microsoft
Ruston Panabaker Microsoft
Mark Overby NVIDIA
Karen Zelenko Phoenix Technologies
Joe Chen Samsung
Marc Noblitt Seagate
Jim Hatfield Seagate
Diepvu Le Silicon Image
Aaron Wilson ST Microelectronics
Yutaka Arakawa Toshiba
Mladen Luksic Western Digital
Curtis Steven Western Digital
John Masiewicz Western Digital
4 Approval of the minutes from the previous T13 plenary meeting (e05126r0)
Mark Overby moved, Mark Evans seconded, that the minutes be approved. The motion passed with unanimous consent.
5 Document Distribution
No documents were distributed by hard copy. Dan Colegrove provided CD’s with the current copy of material posted to the website.
6 Review of Previous Action Items
1. Mark Overby will construct a project proposal for HBA-2 to incorporate the parallel emulation from ATA/ ATAPI-7. [Completed.]
2. Dan Colegrove will query CFA on the use of SET FEATURES subcommand codes. [Carryover]
3. Dan Colegrove to check with INCITS about the anti-trust implications for the SATA-IO proposal. [Carryover]
4. Mark Evans will research if a separate letter is needed from Maxtor for a Quantum patent related to UDMA that was released for ATA-5 and subsequent project. [Carryover]
5. Curtis Stevens will construct a proposal to clarify the behavior as to what hosts shall expect and devices shall do when a command is not supported and for how to report unsupported commands. [Completed]
6. Dan Colegrove will construct a document to propose how T13 and SATA-IO might work together. [Completed]
7. Curtis Stevens will construct a proposal to recommend how devices report support for the new ATA standards. [Completed]
8. Hard Drive vendors are to find out if their drives spin down when they receive a STANDBY command. [Completed]
9. Curtis Stevens to bring in a proposal to include in the definition of DF, that a device requires a power on reset after reporting device fault. [Completed]
10. Dan Colegrove will forward the DRQ=0 when ERR=1 Technical Report Project proposal to INCITS. [Completed]
11. Hard Drive vendors (particularly Dan Colegrove) are to find out if their drives use the I/O and C/D bits. [Completed]
7 Old Business
7.1 ATA/ATAPI-7: 1532D
7.1.1 INCITS Status
INCITS has published the document as an official standard.
7.1.2 ISO Status
The document has been forwarded to Gary Robinson for international standards publication. Mark Overby agreed, as the IR for T13, to shadow Gary Robinson to follow the progress and learn the process for future standards.
7.1.3 ATA/ATAPI-7 Errata: e05108 [Colegrove]
Dan Colegrove reviewed the current list of ATA/ATAPI-7 errata. John Masiewicz noted that the errata about clause 14.4.8 has not yet been incorporated into the errata document. Dan asked for an action item to make sure this was included in the next revision.
7.2 EDD-3: 1572D
7.2.1 INCITS Status [Stevens]
Curtis Stevens stated that he has received the final document from INCITS and that the standard has been published.
7.3 SCT Technical Report: DT1701
7.3.1 INCITS Status [Stevens]
Curtis Stevens stated that he has received no feedback from the INCITS editors about the technical report.
7.4 HBA Standard: D1510
7.4.1 Errata to HBA Standard (e05114r0) [Overby]
Mark Overby reviewed the errata to the HBA standard. Curtis Stevens asked if the standard was incorrect or if the original material was incorrect. Joseph Chen displayed SFF-8038 which was the original source material for the standard. SFF-8038 defined BAR 4 to be 16 bytes of I/O space. The standard currently defines 32 bytes of I/O space. Therefore the error was introduced in the HBA standard. Curtis accepted this conclusion.
Mark Overby moved, Dan Colegrove seconded, that the errata (e05114r0) be submitted to INCITS for further processing. The motion passed 11 yea, 1 ney, 0 abstentions.
7.5 ATA/ATAPI-8 Projects
7.5.1 ATA-8 Schedule Review
Dan Colegrove reviewed the current schedule for the various ATA-8 projects.
ATA-8 Project Schedule
Project / Proposal Cut Off / Stabilization / Letter Ballot / Forward to INCITSArchitecture / April 2005 / August 2005 / October 2005 / April 2006
Commands / June 2005 / October 2005 / December 2005 / April 2006
Serial / June 2005 / October 2005 / December 2005 / April 2006
Parallel / February 2005 / June 2005 / August 2005 / April 2006
Mark Overby acknowledge that the parallel ATA project is behind schedule and stated he would endeavour to bring the parallel ATA draft to closure as quickly as possible.
7.5.2 AT Attachment 8 - Architecture Model (ATA8-AM)
7.5.2.1 Architecture Model and Diagram [Evans]
Mark Evans reviewed the current changes (by change bar) in revision 0a of his document. Mark took significant feedback that will be incorporated in the next draft.
7.5.3 AT Attachment 8 - Command Set (ATA8-ACS)
7.5.3.1 Download Microcode Enhancements (e04132r0) [Livaccari]
This item was deferred as Steve Livaccari was not present.
7.5.3.2 Secure Erase Freeze Lock Proposal (e04147r0) [Hughes]
Gordon Hughes presented a proposal relating to the use of the security feature set and secure erase for ATA drives. Gordon pointed out that the security feature set has some very good features such as not being able to do a security erase unit command without knowing the password. He also pointed out that security erase unit has the highest forensics security, that there is no data left on the drive left to access. However, he acknowledged that there is vulnerability to malware. Gordon proposed that security erase unit be allowed if the password is set or allow security erase unit if issued with the master password or if the security system has been enabled by setting a user password at maximum security.
A lengthy discussion ensued about the mechanisms for the mechanisms for using security erase, security freeze lock, and other security feature set commands. Discussions also centered around why various upstream components of drives do not issue these commands today. Suggestions were also made about logo’s or other material that would inform people about systems or environments that properly used the security sets. There was division between the members about where the problem should be solved and if this is a continuing problem with security erase unit.
Dan Colegrove asked for a straw poll to indicate support for allowing secure erase with a password after a freeze lock event only if a user password is set. The straw poll was concluded as 0 in favor and 10 opposed.
7.5.3.3 Queuing Control Mechanism (05110r1) [Overby]
Mark Overby presented revision 1 of the queuing control proposal to receive feedback about the proposal. He reviewed the changes about removing the set features requirement as well as using a currently n/a bit for controlling the drive responding with data or notifications for the queued command. A significant discussion was held about why this feature set would be useful and Mark agreed to incorporate more of the motivation into the header of this proposal. Marc Noblitt asked if this proposal was only for parallel devices and current ATA queueing methods. Mark Overby responded that this will encompass NCQ as well as SATA technologies in the next revision.
7.5.3.4 NV Ram Buffer Control (e05106r2) [Obr]
Nathan Obr presented revision 2 of Microsoft’s proposal for incorporation of non-volatile cache technology and commands. Nathan displayed revision 3 of the document which contained formatting changes to follow the ATA-8 formatting style.
Jim Hatfield noted that the spindle state was binary (spun up or down) and questioned if the transitional stated needed to be noted. Nathan noted that he believed that the transitional states were not informative for the purposes of this feature set. They noted that the intent was to know if the spindle was spun down to avoid unnecessary I/O causing the spindle to power up and draw significant power.
Curtis Stevens objection to language in section 1 that seemed to restrict the implementation of the flash to within the device. Nathan stated that his intention was to just provide a set of ATA commands to manage such an NV cache. Nathan agreed to word that section differently. Mark Overby objected to pushing this onto the host. Nathan just wants a ATA command set to manage this and he is not currently concerned about where this cache lives at this point.
Rob Strong started a lengthy discussion on the merits about allowing the drive to reclaim some of the pinned set in order to allow for various scenarios of drive intelligence. Nathan reiterated that the intent was for the host to completely manage the population of the cache. A suggestion was made that it would be possible for a device to have a larger cache than what it reports and can manage the unreported cache for its own operation. Nathan agreed based on feedback about pinning, multiple OS behavior, and other cache management scenarios that he would go back and think about it more.
A further discussion was held about the formats of the lists and requiring that the LBA’s be sent to the device in ascending order when pinning information into the NV cache. Nathan agreed to require that the LBA’s be sent to the device in ascending order.
Mark Overby raised a concern about the alignment of the buffers for the cache misses which would require embedded processors to do multiple memory or bus accesses. Nathan agreed to modify the table to have 8- byte alignment and to update the tables.