RFID (request for implementation description): SBC-4 Remove Unused MMC Restrictions (T10/14-238r0)

Paul Suhler Paul.Suhler at hgst.com
Tue Sep 16 11:21:02 PDT 2014


* From the T10 Reflector (t10 at t10.org), posted by:
* Paul Suhler <Paul.Suhler at hgst.com>
*
Thanks very much, Rob.
So, I should also ask whether anyone has information on whether any CD/DVD
manufacturers want the higher-capacity commands.  If I correctly recall a
discussion with John L., the only activity around MMC-6 is to fix some
existing problems.  No one seems to have an intention of resurrecting the MMC
working group.
If you know of any intention to implement MMC devices with READ (16) or WRITE
(16) - either inside T10 or outside - then please speak up.
Thanks,
Paul
-----Original Message-----
From: Elliott, Robert (Server Storage) [mailto:Elliott at hp.com] 
Sent: Monday, September 15, 2014 5:28 PM
To: Paul Suhler; T10 E-mail Reflector (t10 at t10.org)
Subject: RE: RFID (request for implementation description): SBC-4 Remove
Unused MMC Restrictions (T10/14-238r0)
> * From the T10 Reflector (t10 at t10.org), posted by:
> * Paul Suhler <Paul.Suhler at hgst.com>
>
> I created a proposal for SBC-4 to clean up the "Restricted for MMC-6" 
> bits that we found yesterday in six commands that have never appeared 
> in any MMC
> standard:
> 
>    http://www.t10.org/cgi-bin/ac.pl?t=d&f=14-238r0.pdf
...
> I'd like to make a more precise request in regards to this proposal:	
> Does anyone know of any MMC device that sets the restricted bit in any 
> of the following commands:
> 
> READ(16) byte 14 bit 7
> VERIFY(12) byte 10 bit 7
> VERIFY(16) byte 14 bit 7
> WRITE(16) byte 14 bit 7
> WRITE AND VERIFY(12) byte 10 bit 7
> WRITE AND VERIFY(16) byte 14 bit 7
Those were requested by Keiji Katata, Pioneer and the MMC working group and
first went into sbc2r09 (see CAP minutes 03-047).  
>From my notes at the time...
"MMC has assigned those bits in the CD/DVD command set to support streaming. 
This means:
* on reads, if a bad block is encountered, just return zeros
  and proceed to the next block
* on writes, if a bad block is encountered, just skip it and
  proceed to the next block
This is done on a per-command basis.  It lets multimedia software use:
* non-streaming accesses when accessing important data like
  file directories, firmware, etc. where data integrity is important
* streaming accesses when accessing files containing audio/video data
  streams where performance matters more than integrity
This is very useful for DVD players, DVD recorders, PVRs like ReplayTV and
Tivo, etc.  It could be useful for large video servers as well.
Keiji Katata from Pioneer suggested that SCSI disk drives might want to
implement the same features.  T13 has been adding streaming support to the
ATA command set in ATA/ATAPI-7.
The bits used in MMC-4 for the features will be marked as "restricted" in
SBC-2 so they don't get used for some other purpose.
"
Since optical drive capacity lagged magnetic drive capacity, MMC never added
16-byte CDBs; most ATAPI silicon doesn't support that packet size. They never
picked up VERIFY (12) and WRITE AND VERIFY (12) to support > 64 Ki block
transfer sizes in one command, either.	Nobody has proposed adding the
Streaming bits to the SBC commands (a few similar proposals have been
discussed, but they always want something more complex), so they've just sat
as restricted for 11 years.
---
Rob Elliott    HP Server Storage
*
* 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