Request to SBC, SBC-2 from Fuji

Gerry.Houlder at seagate.com Gerry.Houlder at seagate.com
Thu Dec 19 07:14:26 PST 2002


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

To offer my suggestion about the proposed mode page feature and proposed
streaming feature in CDB:

We have not actually used device specific fields within a CDB or mode page
before (it is usual to have device specific mode page or CDB however), but
CDB and mode page fields can be optional for one device type or interface
protocol but be mandatory for another.

I believe the functions being proposed for MMC-4 can be useful for other
direct access devices as well. For instance, mode page 1 has had a "read
continuous" bit that is intended to specify reduced error recovery time per
block in order to maintain a stream of data (intended for video playback
application where keeping data stream going is more important than having
every block correctly recovered). This feature would be more useful if it
could be specified in the CDB (like the streaming field in the new
proposal) so that application can easily specify data that can use reduced
recovery (even if some of the data might be incorrect) versus data that
must be correctly recovered (or the application must not use it). It is
becoming much more common that both types of data reside on same system
with interleaved requests, so using a mode select command to change between
the different recovery criteria would be very troublesome. When the
commands are queued and might be reordered by the target device, the mode
select technique to change recovery criteria becomes impossible.

Therefore I would support defining the functions (both the CDB field and
mode page field) in a common manner for all direct access devices. Each
device type can then use the fields or not based on their market needs.



                                                                                                      
                    keiji_katata at post.pio                                                             
                    neer.co.jp                   To:     "Elliott, Robert (Server Storage)"           
                    Sent by:                     <Elliott at hp.com>                                     
                    owner-t10 at t10.org            cc:     <t10 at t10.org>, <mtfuji5 at avc-pioneer.com>,    
                    No Phone Info                <twg3 at ml.cds21solutions.org>                         
                    Available                    Subject:     RE: Request to SBC, SBC-2 from Fuji     
                                                                                                      
                    12/18/2002 08:37 PM                                                               
                                                                                                      
                                                                                                      




* From the T10 Reflector (t10 at t10.org), posted by:
* keiji_katata at post.pioneer.co.jp
*

Thank you Elliott,

I forgot to write the most important question. I would like to check that
SPC-x, SBC-x already has the similar function or proposal like "Enhanced
Defect Reporting and EMCDR field", or not. As the result of my personal
survey, I can not find any. In MMC4 meeting, I also do not receive such
comment. If similar function has been defined already, MMC/Fuji people need
to examine to use the function, I think.

As far as I know, Mode Parameter page code and its contents can be device
type dependent. So I think MMC4 can have EMCDR field in MMC4 Read/Write
Error recovery mode page.
I'm not sure that Command Descriptor Block has same manner with Mode
Parameter. In case of Streaming bit in Read 12/Write 12 command, currently
it is MMC device specific bit definition. If this is allowed in SCSI
environment, I need not propose Streaming bit of Verify 10 to SBC, SBC-2.
Actually command description is device type specific now. For example Write
10/12 of MMC has restricted overwrite function. This is completely
different with the Write command in SBC, SBC2. Sequential (e.g. Tape)
device also has its specific definition.

In case of 16 bytes CDB length, ATAPI device can not use it now. Because 16
bytes CDB is defined newly, many software and hardware e.g. tailgate
interface can not accept it.

Anyway, I would like ask 2 questions to all T10 people.

1. Do SPC-x, SBC-x have similar function with "Enhanced Defect Reporting
and EMCDR field"?
2. Is it allowed that CDB has device type specific field?

I will be in MMC4 WG meeting room. If CAP people want to ask us (me) any
question, I would like to visit CAP meeting room. Let me know at the time.

Best regards,

Keiji Katata
PIONEER CORP.


*
* 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