[T10] Operation of TAS bit related to Unit Attention

Ralph Weber roweber at ieee.org
Tue Jan 26 12:55:48 PST 2016


I tend to agree with George.

Originally, the TAS bit definition in SPC held sway. As time went by, 
Rob succeeded in moving the majority of the TAS requirements to SAM. 
Now, SAM repeats the SPC description and greatly elaborates on it.

On the other hand, there is no conflict between the SAM and SPC 
descriptions, unless one has a minimal understanding of what a Unit 
Attention is.

The years also have not been good to the "(see SAM-5)" reference at the 
end of the SPC TAS bit definition. Prior to the invention of the Note to 
Entry, this at-the-end format was intended to mean "everything about 
this definition is clarified in SAM-5". Now, it's a dangling oddity.

With this in mind ...

The minimalist change is to replace the SPC-5 "(see SAM-5)" with a new 
last sentence on the paragraph along the lines of, "SAM-5 describes the 
effects of the TAS bit in detail." Based on the notes above, I could 
consider this to be an editorial change.

The big-foot change is to remove the "(see SAM-5)" at the end of the SPC 
paragraph; begin the paragraph with "As described in SAM-5, ..."; and 
replace the two instances of "shall be" with "are". This is clearly new 
proposal territory.

All the best,

.Ralph


On 1/26/2016 2:08 PM, Bill Martin-SSI wrote:
>
> The biggest confusion here was the apparent opposite meaning of the 
> bit, but now it is clear that the functionality of the bit was 
> designed in that when set to one tasks are completed with TASK ABORTED 
> status and there is no UA. When set to zero tasks are aborted with no 
> response to the application client and a UA issued.
>
> Thanks,
>
> Bill Martin
>
> Vice Chair SNIA Technical Council
>
> Vice Chair INCITS T10
>
> SSD I/O Standards
>
> Samsung Semiconductor, Inc.
>
> Cell (408) 499-1839
>
> *From:*Black, David [mailto:david.black at emc.com]
> *Sent:* Tuesday, January 26, 2016 10:25 AM
> *To:* Bill Martin-SSI; George.Penokie at avagotech.com; Ralph Weber 
> (roweber at ieee.org) (roweber at ieee.org)
> *Cc:* T10 Reflector
> *Subject:* RE: Operation of TAS bit related to Unit Attention
>
> I’m neither Ralph nor George, but I don’ t see a serious conflict:
>
> - The SAM-5 text is about creating a unit attention condition that is 
> reported on a **subsequent** command (if it is reported, see SAM-5 on 
> UA queuing behavior).
>
> - The SPC-4 text is about how to terminate the **current** command.
>
> It might be nice if each chunk of text acknowledged the existence of 
> the other, but that would be editorial.  What have I missed?
>
> Thanks, --David
>
> *From:*t10-bounces at t10.org <mailto:t10-bounces at t10.org> 
> [mailto:t10-bounces at t10.org] *On Behalf Of *Bill Martin-SSI
> *Sent:* Tuesday, January 26, 2016 11:11 AM
> *To:* George.Penokie at avagotech.com 
> <mailto:George.Penokie at avagotech.com>; Ralph Weber (roweber at ieee.org 
> <mailto:roweber at ieee.org>) (roweber at ieee.org <mailto:roweber at ieee.org>)
> *Cc:* T10 Reflector; Stefanus (Fnu) Stefanus-SSI
> *Subject:* [T10] Operation of TAS bit related to Unit Attention
>
> Ralph & George:
>
> There is a conflict between SAM-5 and SPC-4 on the meaning of the TAS bit.
>
> SAM-5r21 states:
>
> If the TAS bit is set to zero in the Control mode page (see SPC-4), 
> then the device server creates this unit attention condition for each 
> I_T nexus that had command(s) aborted other than the I_T nexus that 
> delivered the command. If the TAS bit is set to one in the Control 
> mode page (see SPC-4), then the device server does not create this 
> unit attention condition.
>
> SPC-4r08 states:
>
> A task aborted status (TAS) bit set to zero specifies that aborted 
> commands shall be terminated by the device server without any response 
> to the application client. A TAS bit set to one specifies that 
> commands aborted by the actions of an I_T nexus other than the I_T 
> nexus on which the command was received shall be completed with TASK 
> ABORTED status (see SAM-5).
>
> Which was the intended meaning of this bit?
>
> Thanks,
>
> Bill Martin
>
> Vice Chair SNIA Technical Council
>
> Vice Chair INCITS T10
>
> SSD I/O Standards
>
> Samsung Semiconductor, Inc.
>
> Cell (408) 499-1839
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.t10.org/pipermail/t10/attachments/20160126/97293120/attachment.html>


More information about the T10 mailing list