From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Sheffield, Robert L
Sent: Thursday, December 02, 2004 2:57 PM
To: T10 Reflector
Subject: SAT: Notes from the Dec-2 '04 SAT WG Teleconference.
SAT WG Webex/Teleconference
December 2, 2004 9:00 AM MST - 12:00 PM MST
Attendees Logged (perhaps incomplete):
Bob Snively - Brocade
Kevin Marks - Dell
Curtis Nottberg - Emulex Longmont
Ralph Weber - ENDL
Wayne Bellamy -
HP
Rob Elliott - HP
Bob Sheffield - Intel
Nate Marushak - Intel
Mark Evans -
Maxtor
John Lohmeyer - LSI
Mark Overby - Nvidia
Jim Hatfield - Seagate
1. Approval of Agenda
2. Attendance and Membership
3. Old Business
3.1 SAT INQUIRY contents (04-218r6) [Elliott]
3.2 ATA Command Pass-Through
(04-262r6) [Stevens]
3.3 SAT:
Remaining Work Items (04-391r1) [Sheffield]
4. Review of SAT Working Draft (SAT-r01.pdf)
Rob went through his latest INQUIRY proposal.
Decided to keep the "shall implement this field as
defined in SPC-3.."
that Rob had used to
replace text for several INQUIRY field descriptions.
Questions about BQUE and firewire.
SAT currently says it shall be 0, consistent with what Ralph
had lobbied
for in previous SAT WG
meetings. Rob's proposal says "see SPC-3". Fixing
BQUE at 0 excludes 1394/firewire which says it shall be
1.
Agreed to defer to SAM-3/SPC-3 and add
informative/normative annexes if
needed to
cover transport-specific constraints or recommendations.
Debated whether the logical unit name derived from
model number and serial
number ID should be
ASCII or UTF-8 code set. The group reaffirmed that ATA
is required to report only "printable" ASCII, and so the
group agreed to
specify on ASCII. It's the
device's problem if it doesn't comply - SATL
won't fix it up.
The group did not review Curtis Steven's ATA Passthrough proposal.
Bob Sheffield reviewed the not-yet-posted version
of SAT Work Items
(04-391r1). This didn't
generate any notable discussion. Bob added a note
to add a work item to add SAT draft clauses to address
transport-specific
considerations of
SAT.
Bob Sheffield walked through interim revisions
being made to the working
draft. Made
some editorial corrections. Simplified "e.g." text on new UNIT
ATTENTION subclause. Reviewed adding support for
descriptor-format sense
data - needed to
report sense for 48-bit LBAs.
The group discussed the addition of MICROCODE
DOWNLOAD support in WRITE
BUFFER at length.
Decided modes 6 and 7 (as opposed to modes 4 & 5) are
required since most applications download microcode in
multiple transfers.
The SATL needs to
accumulate the buffers and send one complete image to the
ATA device. Need to show interaction between CDB field
options and the
supported modes (editor's
note added).
