DS bit usage in log page 00

Gerry.Houlder at seagate.com Gerry.Houlder at seagate.com
Mon Jun 22 11:21:40 PDT 2009


* From the T10 Reflector (t10 at t10.org), posted by:
* Gerry.Houlder at seagate.com
*
So this feedback from Kevin is along the lines of what I fear could be
abackwards compatibility problem for many customers. Moving the DS bit to a
reserved bit in the page code byte of returned log page data could break a
lot of system utilities that expect that reserved bit to be set to zero or
the utility will post an error. Log Page zero is slightly more significant
because if that one fails, then the utility may not even try other log
pages. Of course, if it does try other log pages then they may also fail
for the same reason if the DS bit is set.
Is this enough of a problem that the DS bit should be moved back to the
parameter flags? Or should we simply define log page zero (and no other
pages) as NOT having a DS bit (i.e., let that bit remain reserved)? Any of
these changes will require a proposal to be written. I am willing to write
it if I think there is enough guidance from reflector traffic as to the way
it should be written.
	     Kevin D Butt						   
	     <kdbutt at us.ibm.co						   
	     m> 							To 
	     No Phone Info	       Gerry.Houlder at seagate.com	   
	     Available							cc 
				       owner-t10 at t10.org, t10 at t10.org	   
								   Subject 
	     06/22/2009 12:32	       RE: DS bit usage in log page 00	   
	     PM 							   
If you change this you are begging for most, if not all, of your legacy
applications to break.	IBM will be voting for supported pages to always
return 0.
Kevin D. Butt
SCSI & Fibre Channel Architect, Tape Firmware
MS 6TYA, 9000 S. Rita Rd., Tucson, AZ 85744
Tel: 520-799-5280
Fax: 520-799-2723 (T/L:321)
Email address: kdbutt at us.ibm.com
http://www-03.ibm.com/servers/storage/
 From:		  Gerry.Houlder at seagate.com				   
 To:		  t10 at t10.org						   
 Date:		  06/19/2009 02:25 PM					   
 Subject:	  RE: DS bit usage in log page 00			   
* From the T10 Reflector (t10 at t10.org), posted by:
* Gerry.Houlder at seagate.com
*
I am posting this to SCSI reflector because it turns out that moving the DS
bit from the parameter control byte to the page code byte may cause
backwards compatibility issues for some companies. The example discussed
thi these emails is the Supported Log Pages page, where the page code byte
would change from 0x00 to 0x80 if the DS bit was set for this page. Setting
this bit seems like the right thing to do for this page (it is described as
being undefined for LOG SELECT, so it seems not to qualify to have any
savable parameters) but software that tries to retrieve this page may be
confused by the result.
Pages that have savable parameters should not see an issue (the DS bit = 0,
so will look the same as before).
I ask that companies evaluate whether this item will be a problem for their
software that reads up log pages. I would like to hear opinions at the July
T10 meeting from companies on this issue.
	    "Elliott, Robert
	    (Server Storage)"
	    <Elliott at hp.com>					       To
	    No Phone Info	      "Gerry.Houlder at seagate.com"
	    Available		      <Gerry.Houlder at seagate.com>,
				      "roweber at ieee.org"
				      <roweber at ieee.org>, Mark Evans
	    06/19/2009 02:49	      <Mark.Evans at wdc.com>
	    PM							       cc
								  Subject
				      RE: DS bit usage in log page 00
The history is:
1. Before SPC-3 letter ballot, log page byte 0 contained an 8-bit the PAGE
CODE field
2. In letter ballot, byte 0 bits 7:6 changed to Reserved per a Dell LB
comment that the CDBs only had 6-bit PAGE CODE fields.
3. At that time, each log parameter had its own DS bit
4. 05-242r2 (by Mark) made the parameter-specific bit DS bit obsolete and
created a new page-specific DS bit in byte 0 bit 7
The Supported Log Page page did not have conventional "log parameters" so
never had DS bits.
I think the Supported Log Pages log page (00h/00h) should always return
DS=0 to remain the same as it has always been.	The Supported Log Pages And
Subpages log page (00h/FFh) and Supported Subpages (xxh/FFh) log pages
should follow suit.
> -----Original Message-----
> From: Gerry.Houlder at seagate.com [mailto:Gerry.Houlder at seagate.com]
> Sent: Thursday, June 18, 2009 7:52 PM
> To: roweber at ieee.org; Mark Evans; Elliott, Robert (Server Storage)
> Subject: DS bit usage in log page 00
>
>
> Hi gang,
>
> I have noticed that log page 00 (the Supported Pages log page) has a
> statement that this log page is not defined for LOG SELECT, only for
> LOG
> SENSE. Does this suggest that the DS bit for this page shall be set to
> 1?
> The definition for this bit suggests that this bit is mostly useful for
> LOG
> SELECT. Should something be added to the log page 00 description
> defining
> the DS bit as set to 1 for this page?
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org
*
* 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